Coveo Solutions Inc. (CVO) Earnings Call Transcript & Summary

Good morning to those in the room and online via the webcast, and welcome to Coveo's Second Annual Capital Markets Day. My name is Paul Moon, I'm the Head of Investor Relations. This event is being recorded, and a replay will be accessible on our IR website at the conclusion. A big shout out to the TMX center, market center, I should say, for this lovely room. As you may know, there is typically an opening bell ceremony happening next door. So just be aware of some of the noise that may be associated with that at around 9:30 just so that heads up. So we've got a great line of speakers today. You're going to be hearing from these folks, and we're really happy with our panel. We've got a couple of great speakers. We've got Suzanne Krpata, she's the Chief Operating Officer at SAP, CX. And we've got Prithvi Mulchandani, Vice President of IT Business Applications at Deltek. So thank you so much for coming to this event. So you've had a bit of time with agenda, so I won't review this. You can see that we've got a little bit of a break built in. We've got a Q&A session as well. And as you can see, we've got a standing mic. I would ask that if you have questions, please wait until this session, and we'll get your questions answered. Also for those on the webcast, there is a dialogue box, feel free to use that to submit your questions. And if we cannot address all questions due to timing, we'll be sure to get back to you offline. Just quickly, any financial figures that we present will be in U.S. funds unless otherwise stated. This material should be considered alongside of the continuous disclosure documents that we filed, including our AIF and the most recently financial statements. We will be making references to forward-looking information, IFRS, non-IFRS metrics, KPIs used in our industry and a definition of those are available in our continuous fiscal documents. So without further delay, I would be pleased to introduce Louis Tetu to the stage, Louis?

Thank you. And I think I'll grab a mic or pick this one? All right. Can you hear me? Very good. So let me try to get the clicker to work here. Okay. Perfect. So first of all, welcome, everyone. We really appreciate the time today that off your busy schedules. And so delighted with this day and hopefully, you find it worthwhile. We're very excited, obviously, with the industry, with the company and with essentially being part of that AI revolution and its application within business. And this is what I want to talk about is give you some context about the why and about what is it that we do. And then afterwards, we'll have some conversations with SAP and Deltek, which will make it very, very, very practical. But we're part of really, really an important revolution. And obviously, there are a number of players in this space, and we'll try to position that for you and really explain why we think Coveo offers something that is highly, highly valuable and potentially is a highly valuable asset. And this really is all about the -- what we call the trusted AI experience advantage and why that matters because as for any company, it's all about the why. And why would anybody care about what we do and what is it that we do that is so unique. Coveo has been in AI for more than a decade, and building its platform essentially with large enterprises. So we understand what it takes to build what we call a trusted AI advantage. We built, again, one single platform working with leading brands across the world. And as we'll see, and I'll get a little technical, not too long. I understand this is a financial audience, but just to position things today. We'll talk about how we can deliver in a very unique way, the combination of semantic search, AI-powered recommendation engines, now generative answering as well and high degrees of personalization that really are critical for enterprises. And this essentially is what we've been doing for 10-plus years. So essentially, where we stand today is really the cumulative work of 10-plus years of working with these large global companies. And we've gained a lot of recognition, obviously, leader for more than 5 to 7 years in a row with Gartner, Forrester, high degrees of SLAs which is service level agreements. Basically, we operate at Five9, where we obviously have all the compliance to basically deal with large organizations. So bottom line is we can scale. You're going to hear from Suzanne at SAP today. We're #1 partner of SAP in the world in that area, one of the top partners at Salesforce, we're a gold partner of Adobe. So we're up there basically highly, highly recognized. But again, this is because over all these years, we've been evolving our platform. AI is not new to us. The best thing that happened to us was obviously ChatGPT. And the reason is not so much that it generative. We're going to talk a lot about generative today. But the reason is that it woke up the world to the idea that AI creates a different paradigm in digital experiences. And so I always joke and say it took -- it took people -- it took chat GPT and people writing poems on an iPhone in January and last December to only realize that AI is an extremely powerful technology. Obviously, we knew that. When I joined Laurent in 2008, he was already talking about how Jeff Bezos was using AI at Amazon, and how Netflix was using AI to personalize your experience. So this is exactly what we're going to be talking about today. So our journey started in 2010 when we were essentially a Gartner top right already in search engines. And we started taking that stack to the cloud and applying behavioral machine learning and our journey took us through deep learning, semantic search. We started using large language models in 2020. We released a technology named Smart Snippets worked with a number of our clients to deliver essentially exerts of content as opposed to links to documents is really, really critical and digital experiences. And our journey doesn't stop here essentially. Now we will show you today generative answering, we are -- we believe we are the first globally, the first company to go -- to have gone globally in full production with generative answering in real-time digital experiences and we'll show you examples of that today. And our SaaS model is really a subscription for our customers, is really a subscription to ongoing innovation. People think about SaaS, SaaS is a delivery model, but not all SaaS is equal. Some SaaS companies deliver software through a subscription and charge on a monthly basis, but it's the same software. In our case, it's a subscription to innovation because we update our software 20 times a day to all of our customers simultaneously worldwide. So when customers subscribe to us, they subscribe to our ongoing stream of innovation in AI and we have about approximately 300 people, data scientists and developers improving that platform all day, every day. And I don't think that's something that's well understood in the industry, but it's certainly well understood by our customers. We deal with large, very large enterprise clients that have massive amount of data and massive audiences. I'm going to talk about that, about 675 global enterprises, large or global enterprises use our technology across multiple sectors, such as the tech sector, manufacturing, retail, health care, the financial services sector and many other sectors as well. Fundamentally, what we do is moving companies, helping companies move from persona to persons. So if you think about the way companies operate with -- think about a website, we're going to get into that in a few minutes. Essentially, companies work with data models and intend to put people in specific categories. What AI allows you to do is not to manage 15 or 20 persona. What AI allows you to do is to deliver 1 million different experiences to 1 million different persons oftentimes at the same time. And so this is extremely powerful. And this is something that is not humanly possible without AI, as we're going to talk about. So this ability to curate highly individualized experiences that are prescriptive in nature that are optimized for outcomes. The analogy I've given to many of you in conversations is think about your Netflix experience when you log on to Netflix, Netflix asks the most important question, who's watching. And once they know who's watching, they assemble using machine learning, basically using the sum total of data of all the other users as well as what they can observe from your own behavior. What you browse, what you looked at, the trailers you looked at, what you dropped, et cetera. And based on that in real time within 10 millisecond, they assemble an experience for you. You can search Meryl Streep, you don't need to. All you need to do is browse and press play. That experience is optimized using AI and is designed to delight you to create a formidable experience for you, but it's designed for them also to maximize you're viewing time and maximize their benefit. Think about that same concept in retail. Think about that same concept in the area of customer experiences and et cetera. So only AI can allow you to manage to essentially deliver within milliseconds to millions of individuals at billions of moments at scale. And this is why this technology is so important. So what we've created is a platform that we now call the composable AI search and generative experience platform, and it's powered by a set of relevant AI models that we've built, that are very mature AI models that we've built over more than a decade that essentially support the full cycle of digital experiences across an enterprise. So think about any enterprise an abstract for a moment, whether you deal with a retailer, a distributor, a manufacturer, a health care organization, a bank, you're going to interact through a website, you're often going to go into some sort of transactional site. We're going to hear from Suzanne today, SAP is the largest in the world by gross merchandise value in digital commerce. And Suzanne, the COO, is going to be with us today. And then oftentimes, once you buy something, you want to come back, you want to come back for the warranty, you want to come back for customer service. Maybe you're going to try to self-serve or talk to an agent. And then you've got people within organizations that interact through digital experiences, the employee experiences through workplace applications, internet, et cetera. We've built a platform that essentially supports those areas. And so if you think about the importance of this, it's really, really important. The application of AI within an area such as commerce, increases. And those are metrics that we've actually measured with customers because we're in an interesting world. We're in a world where we can do A/B testing. We can deploy our technology, route part of the traffic to the legacy technology, route part of the traffic to hours and compare in the area of commerce, it's about conversion rates and card side. Ultimately, we've already published that. We're working on algorithms that will optimize margins. How do I create a delightful experience for you online that at the same time, can maximize margins? That's really interesting. And so things like increasing double-digit revenue, revenue per visit or conversion rate. In the area of customer service, it's all about self-service. People prefer to self-serve. I'm yet to meet the person who wakes up in the morning and says, "Boy, I really want to talk to someone in a contact center today" this is not experienced people like. And when you do talk to an agent, you want that agent to have the right information as well and have the kind of the knowledge that will make that agent more proficiency to resolve their issue. We also do that. So we push knowledge in highly contextualized ways and on customer communities, customer self-service sites, et cetera, and we can increase self-service, increase call deflection, decrease the number of agents and make customers happier at the same time. And this is, again, the power of AI. In the area of websites, obviously, it's about increasing conversion time on site and getting people what they want. And in the workplace, which is often overlooked, frankly, it's about doing better than SharePoint. It's about giving people the information they need to get more proficient with their work. Think about the ability of people to do more on their own essentially, through the power of providing them highly individualized information and content in the course of work. And so our platform essentially supports that, and this is a platform that's been 10 years in the making. Today, just to give you a sense of scale, we operate basically across North America, Europe and parts of Asia Pacific with our data infrastructure, pardon me, and handle on a monthly basis, a little more than 25 billion events for our customers. We're talking about billions of searches, billions of index updates and more importantly, we update these customers, again, on a monthly basis, 1,500 times. So we push innovation on an ongoing basis. As I said, this is about subscription to innovation. And we now have the ability across North America through something called Active-Active to operate at Five9s, which is 99.999% service levels with all the certifications and so on. So we're truly enterprise great. Why does that matter? It matters, and I kind of talked through that. It matters because experience is today's competitive front line. everybody, everybody talks about experience. This is now a board-level conversation. I can say that I'm on the board of Couche-Tard and experience -- customer experience is a topic. It's clearly a topic. You have to deliver a better experience nowadays. Products and services, frankly, are a commodity. It's a -- you can get products by clicking plus in your browser, getting in a window and getting to a product or a new provider or so on. Service and your experience is key. What people expect online are experiences that, first of all, are just for me, our design are highly, highly individualized. I no longer expect a vanilla experience. They expect experience is better than that, that can anticipate that are prescriptive in nature that can actually anticipate their needs and make intelligent recommendations, and they expect experiences that are coherent. There's nothing we all hate more than talking to a telco, for instance, about our TV or mobile device and be transferred to another department and be transferred to another department or my favorite one is we'll transfer you to the retention department. No, I just want to cancel my plan because I don't -- I'm not getting great service. And this is something that nowadays is so critical that people don't tolerate. And really, the reason is that choice is a commodity. That when you're online, the one thing that we all have is choice, in fact. And so the volatility of goodwill is greater than ever nowadays. So that's really, really important. However, these experiences are not only individualized, relevant, prescriptive, et cetera, they need to -- companies need to make money. So you need to make sure that you can deliver experiences that drive profitability. I'll give you a simple example of that in retail. If I want to give a great retail experience to consumers, all I have to do is push them popular products at 50% discount. Consumers will be absolutely delighted. The only problem is the retailer will go bankrupt. And that's what's happening right now. In fact, without realizing it, that technology, in fact, online and retail sites, in particular, is driving companies down the tube. And that's because -- and that's a problem AI can solve because AI can figure out that someone is a candidate for buying a high-margin product, in fact, and we'll be just as delighted and the company can make money. And while another consumer may be a little more price sensitive. So the winning experiences, it's already proven, there's multiple sources here. But from a customer long-term value perspective, valuation multiples, brand of choice improvements, et cetera, it's already proven that AI is driving better experiences and is driving -- is driving much superior value for these companies. And there are multiple other sources as well that we can provide. The problem -- so we all agree, every company agrees that creating better individualized experiences is important. The problem, if you don't have AI, if you try to do that manually, the problem is the following. It's sort of -- it's what we call the no win trade-off. If you essentially try to deliver more generalized experiences, you kill the experience and people will go elsewhere. If you try to deliver superior individualized experiences without technology, you're going to bust the budget. If you don't know neither, you're going to go out of business. AI is the only way to solve both. Think about a store, you can create more personalized experiences in a store by adding more clerks. Ultimately, you're going to go bankrupt. It's going to cost too much. If you remove too many -- if you don't have any clerks on the shop floor, it's going to be about experience. So think about that same analog example in the digital world essentially. You need to do both, and that's what we call the AI experience advantage. The power of AI breaks the trade-off, the common attribute between all of our customers, whether they're health care, they are large retailers, large distributors, health manufacturers or financial institutions. They all have the following in common. They have massive amount of content. Lots -- high variety and volume of content or products. We have retailers that have almost 1 million SKUs, and we have retailers that have 1 million consumers a day. In fact, during the pandemic, that's what -- and we're going to get through the Black Friday now. Some of our implementations have 1 million visitors a day. And so massive amount of content or products from multiple sources that need to be delivered to people with high degrees of personalization expectations and a very large and diversified audience often time across the world. So that's a key challenge, especially if you want to make money. And so only AI can solve that problem. In fact, there is no way to solve that problem at scale without using artificial intelligence. Now generative AI brings another layer to this, which is quite interesting because generative AI creates a new form of digital experience. Prior to AI, the way I often explain AI is prior to AI, software was about writing rules on data models. You wanted to write any piece of software, you created a data model and you wrote rules, business rules. And software was mostly about automation and gaining efficiency. Efficiency is doing more of the same. AI is about proficiency and augmentation. Proficiency is not about doing more of the same. It's about enabling people -- augmenting people to do more on their own. And from that perspective, it's -- this is really, really important. Think about a bank. Think about financial advisers in the bank. How could you use AI in fact, so that they could do more on their own without escalating. They could handle more sophisticated problem. How do I activate a mortgage for a commercial property in Fort McMurray with environmental issues and get that answer like that. And that's what we're talking about. And this is really, really important. If you can get in a bank tens of thousands of employees, for instance, to do to handle 10% more complexity, 10% of the time, 10% more of the time. You just created billions of dollars of value. And this is what we're talking about, and we're going to demonstrate some of that today. If you think about -- if you think about the evolution of the digital experience, I'm old enough and maybe some people in the room have known the vanilla website. 25 years ago, you went to a website, you and I went to that website, wherever we were in the world, it was the same thing. It was vanilla. And then some degree of personalization started detecting, okay, your URLs from Germany, I'll give you a German website. So that was the first level of personalization and so on. And then the search box appear you could type your intent in that search box, and it was keyword search initially. And then federated search and contextual relevance. So you started using a little bit of data to understand who was at the other end and contextualize a little bit the answers. AI was the game changer because circa 2011, 2012, AI allowed us to understand who's at the other hand, start observing the behavior really triangulate the context, the intent, the behavior. Becoming more prescriptive with recommendations, and that's when we saw the appearance of recommendation engines on websites, which is, hey, we think you're interested in this. We think you would like that. your peers also do this, et cetera, et cetera. That's about 10, 15 years ago. With generative AI, you're getting into a whole new world of advisory, you can ask generative AI. What is the difference between A, B, C and in what circumstances should I use B? That's pretty interesting. And that fundamentally becomes advisory, and it changes the entire experience. So basically, our point of view is AI or die. If you're a large business, and this is why this matters so much. The quantum leap in experience is so important that to us, it's AI or die. You're either going to adopt these types of technologies if you're an enterprise or you're going to implicitly or maybe unconsciously elect to compete against it. And frankly, if you're in the latter category, give me your stock ticker I'll short you because you're going to lose. You're going to lose doing that. So for us, the future is business to person and that business, the person business-to-person singular, bring the entire business to every person at every moment. And for that, enterprises need that spinal ability to power these trusted experiences, and that spinal ability is powered by AI search and generative experiences. There's just no way around it. Whether it comes from Coveo or from anybody else, there's absolutely no way about it. And fundamentally, if you think about that experience, it's powered by the search box. But you won't have -- the future is converging. Think about -- you're not going to go on a digital experience, whether I'm abstracting here, whether it's a website, a commerce site, the customer service side and Internet, you're not going to go on a page where you have a search box that allows you to navigate results. You have a question box where you can get answers and then you have a chat bot pop on the bottom. This world is all converging. It's an intent box. The world of search, discovery, recommendations, personalization, chat conversations are all converging into one single paradigm. This is the platform that Coveo has been building. And I've already talked about that. So it's the combination, in fact, of these abilities and Laurent will demonstrate some of that. Now in introducing what Laurent is going to talk about, I'm going to be technical for 2 minutes for those of you, and I met people this morning. I met someone who said I was an engineer at Parametric Technology. So -- so this -- because this is important because this is what's under the hood, but I'll try to abstract it to simple things. When you look at large enterprises, they have engagement apps layer. And through these engagement apps, they take people through digital journeys. And they're all different. So think about your own experience as a consumer, you interact with Royal Bank or you interact with Canadian Tire or you interact with any company or maybe you try to interact with the government. And at least in Canada, you need to work hard for that. And so you go on a website and then you're going to go through various properties, right, and through your journey. And each person is different and each journey is different. So this is what our clients have. And so your websites can be powered by Adobe or Sitecore or Acquia or a mix of all the web content management you may have different platforms for commerce, you may have different venues and platforms for customer service, et cetera. And we all know that we all have a lot of logins into the workplace into a bunch of applications. At the lower level, there is a content and data layer. So there's CRM content, there's ERP content, financial data, data records, documents everywhere in thousands of places oftentimes. What Coveo does essentially is it connects these 2 worlds. And so what we provide is 2 abilities. First of all, a spinal ability to unify. It's called indexing. We didn't invent it. Think about Google. Google is an indexer by the way. It indexes the web. We do the same thing. It's called search engine. We do the same thing within secure content within an enterprise. And so we provide the ability to unify that data, we crawl that data with the security. And then at the top, we provide the spinal AI ability for companies to deliver highly individualized intelligence across -- within basically any digital experience. And that's what we call the relevance layer. So that's my 30 seconds of technology here. So we have essentially very extensive plumbing, yes, plumbing that we've built over 15 years, to essentially connect to content at high levels of performance and obviously, uncompromised security. And that's really, really important. So think about hundreds of millions of documents across a global company in 40 different languages across 20 different sources of content combined with real-time data. That's what we can handle. So we can crawl that content. We've been in the search engine business for more than 15 years, and we're very, very good at that. What we've added, in fact, over the past year is the ability to inject in that index, what we call embeddings. And I think there's an [ IP ] on the other side. In beddings and product vectors and et cetera. And we've built AI models. So we have -- the Coveo platform offers to customers a combination of AI models, behavioral AI models, deep learning models that detect intent and so on. And now models that rely on large language models and generated models that can be combined essentially to create relevance, what we call relevance. What is the relevant content and recommendation that I will deliver to David now within that digital experience. And of course, we have all the integration layers. In fact, that allows us to integrate within the most popular apps. We have the analytics. We have the admin tools, and we have extensions such as merchandising intelligence and so on. So this is fundamentally the platform that we're offering. So Coveo is sort of -- I hate to call it a middleware, but this is really an intelligence layer for companies. And so it's very flexible. So whenever we attack, we tackle a website use case, for instance, we can invoke some of these models. In commerce, obviously, we need models around session-based product recommendation or what we call intent to wear ranking. So give us 3 clicks. We will change the ranking and the order of things. We will -- for the same query, we understand that you will likely want something else. And we do that -- we can do that in session just by looking at behavior of users, for instance, in customer service. We have models around case at classification. Obviously, generative answering will show is a key one. And then, of course, in workplace. So we offer the most flexible platform in the industry and enterprises and CIOs really like that because it creates a lot of flexibility for them for their solutions. So where do we fit in the AI stack? If you think about the whole AI stack, you've got NVIDIA and at the bottom, obviously, AMD, Intel and et cetera, providing the compute silicon. Then on top of that, you've got the Hyperscalers, right? So AWS, Azure, the cloud providers, et cetera, providing these services. Then you've got the large language models. And through out that, we have a great company here in Toronto Cohere, in fact, that provides large language models. So they provide essentially the capability to process language and provide answers. Coveo sits on top of that. We consume open AI cohere and the rest. We're sort of agnostic to them to provide an answer. And of course, on top of that, you've got the stack of apps where people engage. So this is exactly where Coveo fit, and we think it's creating a category, which is I'll refer to as the real-time intelligence layer essentially for companies that spinal ability that I described earlier. And we think this is going to become a really, really important category for enterprises, and we can see that because the message and the positioning of Coveo really resonates. So for us, that creates a big opportunity because if you think about how we operate, we have 1 platform. Our company is divided into 2 lines of business. So we have essentially the commerce line of business and what we call the service and knowledge line of business. And then on top of that, we have vertical teams. So the way we go to market essentially is by verticals. So in commerce, there's a big difference between business-to-business commerce. So think about Caterpillar tractor offering aftermarket parts as opposed to Louis Vuitton, for instance. So the difference between B2B retail and brands and so on versus B2B -- versus distribution, manufacturing and NCPG. And then we have service and knowledge teams essentially across the high-tech, very focused teams on the high-tech sector. We're very popular in high tech, financial services and then all sorts of other enterprises from pharma to large services organizations, health and et cetera. So a couple of takeaways here is what our platform offers is the ability for us to tackle multiple verticals, and we can keep adding that. So that's a key growth lever for us to go after those verticals. But also once we land an account in a core use case, as you can see here, we actually have the ability to expand that account and manage that account and really expand the use of the platform across multiple use cases to consolidate the digital journey. So for us, landing a customer and Brandon will talk about that creates a big opportunity for growth. And the platform flexibility a huge opportunity for growth as well across many verticals, and of course, across the world. The way we engage with customers is not only through technology, we're not just a tech provider. We have partners across the world. From Accenture to Deloitte and E&Y and a number of others. But initially, when we engage, we really take the customer by the hand. We can actually run some tests, some pilots, do business value assessments, evaluate the financial potential essentially of the solution, then we deploy the technology. And once the technology is deployed, we now have all the analytics, the A/B testing abilities and et cetera, to tune essentially the AI models and continue the optimization. So let's say, we deploy with a retail, we can get a retailer up and running within a matter of weeks or a few months, depending on the size. And then once the platform is in place, we can tune these AI models and continue to improve on the analytics and the financial results. And that, for us, creates a big opportunity for value creation as well that we intend to leverage as we evolve our pricing methods and so on. So this becomes -- Coveo really becomes not just a software that you implement, but as I said, a constant stream of innovation in a constant stream of improvements of financial metrics associated with the platform, and we think that's a pretty powerful model. So in Gen AI, Laurent will talk about that. But I want to focus on the fact that Gen AI here, our business in Gen AI is not about code creation or content creation. The specificity of Coveo is about applying generative AI, specifically in the area of real-time digital experiences. Frankly, it's harder. You can generate an image. You can generate content. You can summarize an e-mail and et cetera for our clients and so on. That's all good. Those are great applications of generative AI. What we're talking about here is within 10 millisecond essentially applying, given answers to customers that are accurate, that are secure and et cetera. And this is really what CIOs are going through. And the problem we've solved that Laurent will talk about is really addressing. The reason why we feel very good about where we're standing is because we've solved the key problems associated with the application of generative AI within digital experiences. As I said earlier, it's nice to write a poem on an iPhone. And as you know, these ChatGPT hallucinates. Now if you're the Royal Bank of Canada, or if you're Boeing try to answer a Singapore Airlines engineer with a grounded Dreamliner in Malaysia with a broken engine, you can't hallucinate. And so really, the idea is how do you generate accurate answers from secure content, content that is current in real time that is fully traceable in a cost-effective way. And this is essentially the problem we solve. And the reason we solve that problem is because we integrated both the worlds of search and question answering into one. We essentially resolved the problem associated with factual information working from secure content, different set of facts, et cetera, and we were able to combine what we call extraction embedding and vector technology with an index and with a stock of software that already understands relevance, essentially understands who's at the other end. So we can generate accurate answers to questions like how do I add a new bank account feed for a supplier in Norway, for instance, with information that is secure, that is accurate and current. And again, Laurent will talk about that. So some glimpse of that. And as you will see, as we can generate, we've had -- we're Customer Xero for this. We've had this on our own site for a little while, is we have the ability to generate very sophisticated answers to very sophisticated questions. And again, with the full connection to the source of truth. The first implementation of this to our knowledge in the world is the global deployment that was announced of Xero software. And Xero went live. So Xero for the record, is a software company with 3.7 million businesses consumer -- customers, sorry, in the world in accounting software. And they were able to deploy soup to nuts within a few weeks, essentially and asked us to go live into production and measure case deflection north of 20% on top of existing Coveo. This in financial terms is worth tens of millions of dollars for that company. To our knowledge, this is the only deployment of a live implementation. We've done a little bit of -- we've done a little bit of test here on public data. RBC is not a client of generative AI. But we essentially took some of their public data and essentially asked questions such as what are the differences between a prepaid card, a normal card and a gift card? What's the minimal age for each? And this is the kind of answer we're getting. And of course, the sensor is grounded into the content here. So that will give you a sense of the type of things that we can do, and this is here a similar example that we've done recently with public Air Canada data. How do I -- what fares and status are eligible for preferred seed selection, boom, boom, boom, you get the answer. So you can see, and I guess an image is worth a thousand words. You can see here the revolution that this creates. So I'll conclude by saying that the reason we feel very good about our position in the market is, first of all, we think we're the only ones right now standing being able to do that because we understand secure content, indexing current content at scale and we understand relevance. And number two is we're really built for where that matters for enterprise scale. And so that puts us in a very, very positive, I would say, position in a market that I will say was a bit stalled because ChatGPT took the world by surprise, and every enterprise sort of paused and was trying to figure out what is the application of generative AI and AI, and where -- what are the benefits and et cetera. Now that market is really, really reopening and Coveo is standing with not hype, but real solutions and real results, and you're going to see that in a minute. So with that, thank you very much.

Thank you, Louis. I'd like to welcome to the stage, Laurent Simoneau. Take it away, Laurent.

Thank you, Paul. Thank you, Louis. Good morning, everyone. It's good to see a lot of familiar faces and person instead of Zoom. Look better in 3-dimensional. Okay. So next 25 minutes or so, we're going to cover recent innovations, where we're going, so we'll give you a sneak peek of some key investments that we're making in pilot with customers. And of course, we will start with demos around Gen AI. So we will split the presentation in 2 big parts. The first one is about our knowledge solutions. So think about what we do with custom -- customer service and also workplace. So -- and the second part will be commerce. So in this case here, what are the key themes, the key priorities for Coveo in the coming years. Number one, it's obviously AI and generative answering. We also have key investments in application and integration. So Salesforce is a key partner of us. SAP is a key partner of us and so on. And then data and connectivity. As Louis said, we are a spinal AI across those multiple touch points across those multiple connectors inside the enterprise, so we need to invest in this. We need it. We think that we need to have this capability to connect to the entire enterprise and scalable, secure and reliable fashion. So I'd like to give an update -- on Gen AI, and let's see if we can switch here. So you can hear me, right?

Yes.

Okay. So I'll start with Customer Xero. So that's [ partner.coveo.com ]. That's our partner portal, where there's a lot of technical documentation. There are online discussions, there are blogs and it's scattered across multiple data sources, multiple systems, okay? And so if you look here, that's live, you can go there. I'm not logged in here, right? So there's a lot of content here. That's highly technical. So our customers will go there, our partners will go there to ask complex questions. But first, they may ask things like Service Cloud here, right? So that's easy. And we hope the Internet gods are with us. Yes. okay? So that's a classic search interface. AI-powered, great results at the bottom based on behavioral data that we've captured over time. You've got facets to the left here where you can slice and dice results, right? So that's a Coveo that is being a leading platform provider over the years around search. But now let's do something that is a little bit more interesting, how does Coveo determine relevance. So that's a question that we get quite often, right? With those results, that are supporting the query from a semantic standpoint. So now we've introduced semantic search under the hood, right? So we ask a question, results will match semantically to query. The best passages of those results, the best exerts from those results will drown the prompt to the large language model that runs on our infrastructure. So we're using OpenAI GPT-3.5 on our own Azure. So we're going to ground the prompt with those best experts from those results and then get an answer from that. By doing this, we're based on fresh content, not on what the model was built on or fine-tuned on. We're built on fresh content. It's personalized. It's based on the relevance also that we provide from a semantic search perspective. And it's there to basically provide the best result coming from all of the universe of data sources that support that. And it's contextual, right? So if I -- how does Coveo leverage permissions? So here, what's interesting is I have a more complex answer which technical people do love, but then they say, "Well, but I'm using Salesforce [indiscernible] here" right? So I just click on the facet to the left. I've contextualized my results set in real time, and here's -- the answer will change. The answer will change because the results do change at the bottom, right? So [indiscernible] and Salesforce remission are not working as expected. That's a support article. That ground -- that grounds the prompt gets a different answer. So let's do something else here. So how to create a partner organization. So I'm a new partner. I want to create a partner or of Coveo. How do I do that? Oops, I don't have an answer. Oh, why is that? Well, because I'm not logged in, right? So if I'm a partner, I will log in. We'll see. Now I've logged in, now I've got this personalized portal, where, obviously, I have access to more content. So Coveo deals with all those permissions and all of those security access. So if I type the same query here, how to create a partner organization, there you go. Right? So now I'm having access to those results in my results set fitting from a semantic standpoint, I'm going to surface that in the prompt in real time and get an answer. So we are not using the large language model here to provide us with facts. We're using the large language model to assemble facts and make and provide a nice answer in English down the road in multiple languages here. Okay? So that's cool, but that's for Coveo. So let's see what customers are doing in production. So we mentioned Xero here, huge Coveo customer. They were already using -- they've been using Coveo for many years across multiple touch points with a lot of value created. So what they have decided to do here is run an A/B test of classic Coveo that is already fine-tuned and optimized and pretty powerful from an AI perspective versus Coveo and generative answering. Let's A/B test it, right? So this is Coveo Live. They went live, they stopped the A/B test. So punch line, the A/B was quite powerful. So they say, let's stop wasting with the old experience. Let's move 100% of the traffic here. And you will have very, very, very interesting examples here, right? How do I update my subscription payment details? Remember, they have 3.7 million companies on Xero, so they cannot answer calls each and every day for things like that. So I've got great search results at the bottom, right? But from those search results, we ground a prompt to the large language model and, quite frankly, very quickly, we get tailored answer. And if the content changes for whatever reason the bottom, they upgrade their product, whatever, of course, the answer will change. Well, let's do another one here. Now to add a credit to an invoice. There you go, see? This is running live for those millions of customers. They love it. They love it so much that -- let me use this. See, so that's Coveo before. Like Xero, that's Coveo after. That's what they measured. So 21% case deflection. That's huge. As Louis said, it's pure hard ROI here. And they love it so much that they share the news to the world. So this is content from Xero actually. And Nigel Piper, who is the Executive General Manager of Customer Experience Xero is happy to talk about this. So they went live with this before a product is formally GA. That's credit cool. First time in the past 15 years that a customer is so happy with the results of the beta, it goes live with the beta at this scale. So we're proud of this one, and that's a blueprint of what we are going to do in the coming weeks and months. Okay. Now a little bit more insight on Xero, where we're going with them, right? So I said earlier on that we are dealing with multiple use cases at Xero, multiple touch points, if you want So right now -- oops. Yes. Sorry, there's a little bit of delay here. So right now, we are dealing with what's called self-service success. Think of the search page on the support portal, right? We're also enabling case deflection here at Xero. So there will be generative answering here. So when you're about to log the case, you have a problem, you're typing your problem. We're going to provide answers in that touch point too, obviously. The next one, and this one is super cool, and I'll show you that. We are embedded into the Xero app that is used by those millions of companies. We're already embedded. So what's coming, we're going to surface answer isn't there, obviously. And then for the agents and to service console, all of this will also be available. So if we look at the -- what we call the end product, we see on the right side, well, in the middle to -- that's the Coveo pop up, if you want, inside the Xero app. So people don't have to leave the Xero app if they have questions to ask. This is coming, and this is coming soon, a matter of weeks. So this search box here that currently provides recommendation to the bottom and search results will also provide answers to questions. Okay? So you have to look at Coveo as a multi-touch, multi-use case platform with spinal AI capability that supports those different personas and those different expectations. All of this with the same infrastructure, the same AI models, purpose built for, in this case, for service. With a very precise ability to measure outcomes with, quite frankly, a lot of success with that. We had conversations at the call here earlier on about OpenAI in the cloud starting to receive some enterprise data to provide answers and so on. And my comment was that our customers don't want to share their data with OpenAI. They don't want to push this data to OpenAI in the cloud. They want this to stay into the enterprise. So that's a slide that we built earlier on with all of those conversations with enterprise companies. So how do we support security and how do we comply with their regulation and internal compliance. So first of all, we do secure content retrieval. So you saw even in the partner community, when I logged in, I got different results. So you need to have very advanced search to not share content to people that they don't have the right to see, right? Dynamic grounding means that we adapt to those results based on the semantic search. We have auditable prompts and responses, so we have a whole set of analytics and back-end systems and tools to support all of that. Xero retention is extremely important. We use a generic version of GPT-3.5. We are not sharing information with Microsoft, we're not sharing information with OpenAI, so we make the call for prompt where you see the results, okay? In data masking, this is in the future, in addition to all of that, we are likely going to do what's called data masking in the prompt. So if there is some piece of information that customers don't want to go through the API calls, we're going to data masking and unmask at retrievable. So this is what we do from a security standpoint. Now let's look at the future. We believe -- let shape that with some of you, our belief is that the so-called search box that we're starting to call the intent box is the universal way for users to interact with information. That's our ability, okay? It's more flexible, it's more scalable. People are used to it. They know what to do with it. It works on a mobile, it works on the desktop, it can be enabled by voice. So we believe it's a universal way to interact with information. So the way we're looking at this is okay. So how can we push this further in terms of interactions, right? So that's what we are currently building and about to pilot with a few customers. It's called conversational search. So one -- in the example here that I'm showing, what is the difference between a personal loan and commercial loan? Yes, of course, we're going to show a great answer, right? That already demonstrated that, that's -- it's almost old news with the references at the bottom and so on. But look at the Ask Follow-up here. This is coming. So the Ask Follow-up allows the users to follow up with the search without entering all the contacts, right? So to go into conversation with search. And the system is also going to share or to suggest queries, additional queries for follow-up based on our AI, based on what has been done before and based on potential outcomes also. So in the example here, so I click on how can I apply? How can I apply, right, to what? Well, applying, see? So there is a different answer here, automatically generated, and at the bottom, do you want to apply for commercial loan or apply for personal loan. This is what we're discussing with customers as we speak. This will likely evolve from a [ UX ] and experience perspective. some elements may come out in production faster than others. But directionally, this is what we are building, okay? And we -- we're not a ChatBot vendor. But we believe that a lot of the interactions that are happening in the ChatBot right now would deserve a more flexible experience, a more scalable experience and an experience that, quite frankly, people love a lot more than a very rigid ChatBot. So that's where we're going. Down the road, we may also inform the chat bot based on what's going there, right? So it's a big deal for us. Okay. Let's switch to commerce. Different bulking. Same platform, same capabilities, but obviously different use cases, different personas. So what do we care about in commerce? We care about merchandising. We care about revenue, but profitability. We'll show you a little bit of generative AI, what we're doing with this. And of course, integrations in commerce are important. And Suzanne from SAP will talk about this great partnership that we're having with SAP, okay? So what do retailers really care about in commerce? Well, they care about serving the shopper's need. You're looking for the good old example that I'm giving always as if you're typing fish on a grocery store website, you want fish. You don't want the cookie, the fish cookie. Even if the margins are better on the fish cooking, people will not convert. So you measure relevance with conversion. But then they also care about business outcomes. As we said, they don't want to lose money online. Well, maybe someone, but it's strongly not scalable, right? So you want to care about business outcomes, increasing revenue per visit, also increasing margins per business. So -- and the balance between both of those elements is the key to successful e-commerce operation, and that's not easy. So some of you are familiar with the merchandising hub. This is coming from our Qubit acquisition. We have evolved this over time. It's not part of Coveo platform. It does all sorts of things, right, deal with campaigns and sites and analytics, personalized content recommendations. It's really designed for the merchandiser. We have been focusing on revenue per visitor for a while now with great success with -- and Tom, I think, has a few slides on that later on. But now what we are getting really into is margins per visitor. So MPV basically. We are in pilot right now with a big retailer, and we will get results out of this in early '24. So at a high level, how do these retailers look at this and what is our impact from Coveo perspective. You can look at -- you can look at the world retailer from those 2 axes, right? They care about conversion, they want people to buy, but they want to extract margins, obviously, from what they bought. Right? So what's our role from a Coveo perspective to play in this? So you have those 4 quadrants. You have the loss leaders at the bottom right. High conversion of margin, you have the profit leaders, the right, so we all love that. The bottom -- on the bottom left, I mean, you have low performing, low margins, right? You have the dogs. And then on the top left corner, we have what's very interesting, the underexposed high-margin products. It is new. It hasn't been marketed properly. It's not visible metadata may or may not be arrived in the catalog, right? So we all -- retailers always have to deal with [ DES ]. Profit leaders, that's easy. On the top left side, that's hard. So what do we do from a Coveo perspective? Number one for those retailers, we are going to highlight high-margin products. That's easy. If there's a high break between 2 products that have the same relevance, we are going to obviously highlight the profit leaders. That's easy because we have margin information now. Then when we deal with a lost leader, we're going to recommend profit leaders, right? Because our recommendation of [ Gordon's ] will make sure that those popular high-margin products will be tied to those lost leaders, and we're providing the tools for the merchandisers to decide that. But then what is really powerful is for those products that are high margin, but are varied, as I said, right, because not that properly -- not popular enough, marketing was not done well. We're going to expose these products into those sessions. And by doing so, we believe we're going to increase the margin per visitor, and to do that, yes, we need to increase the margin per product, but also make sure the conversion remains the same or even is better, okay? So that's what we're building right now with a larger tailer. We're looking forward to the results of this. And this is going to be available from a tooling perspective into the merchandising hub for these retailers in '24. Finally, I'd like to show what we're doing about semantic search and generative AI and commerce. In service, generative AI is pretty obvious. I have a problem here is a solution. In commerce, people want to navigate, do you want to go through the data, go through the catalog, discover things, right? So classic surge, all things in equal is more important in the context of commerce. But how could we include generative AI into those use cases? This is something that we are actually building for a large home improvement retailer in the U.S. And this retailer is -- has a lot of how to buying guys, best practices for home improvement available on their website, right, not just a catalog of things. So this is a demo that we've built for them. And now we are rolling that out in production, hopefully, early '24. Tips to build outside kitchen with a barbecue. What do you do with this, right? So this is the answer from Coveo. Now I want to highlight a few things here. First of all, this is a great answer because that's a personal query actually. So I've been doing my own research and I wanted to see if I were in the U.S., how it would look like. This is generated from the buying guys and the how-to content on their website. Great [ tool ], right? Now from a semantic search perspective, here's what we're starting to do. We are highlighting -- we're highlighting to find the components of that answer, see in blue, that can be used for pop-ups and clicks, and expansion of the experience. But more importantly -- oh, now we're using the answer as a bridge to navigation. From that answer, let's highlight the top categories that may help you. Remember, I said earlier on, in commerce, you want to navigate, you want to slice and dice the content you want to discover. It's again not provide you just with an answer. We want to provide you with choices, what better choices and category pages here, right? So you start with that, and then you have the full experience with the facets of navigation that is obviously driven by AI. So the most important facet, the most important choices will be discovered dramatically by AI and provided there. So again, this is going to evolve, but directionally, this is the opportunity in the context of commerce from Gen AI perspective and we're quite excited about this. Okay? So with that, thank you for your time, and talk to you soon.

I'm pleased to introduce Sheila Morin, she's the Chief Marketing Officer of Coveo. Sheila?

Hello, everyone. Nice to see you again. So not new news to you, we are in a very rapidly growing market. AI, Gen AI, search and knowledge discovery is growing fast. You can see it here, IDC, International Data Corporation is projecting this market to go from 5.5 billion to 21 billion in 2027. Huge opportunity. The other opportunity we have at Coveo is that we estimate the quantity of enterprise that fits in our ideal customer profile to be around 21,000 in the targeted region. So Canada, U.S., Europe, New Zealand, Australia. And we have 675 customers in there, so 3% penetration, again, huge opportunity to grow. So how do we do that? How do we augment market share fast and take a portion of this market fast. So we have 5 growth pillars, and Tom and I will talk about each of them. The first one is clarify and amplify, clarify what we do so that everyone understands it. And then amplify it, so more people know that we exist, what we do and what we can do for them. Then lead on AI and generative answering, of course, leverage our key partnership with ACP with Salesforce, expand geographically, huge opportunity in North America, but we're scratching the surface in Europe and ANZ and accelerate our growth within our own installed base. So let's start with the first one, clarify and amplify. So we are in the back in the growing market, a very noisy market. So a lot of noise, a lot of hype. But how do we cut to the clutter? We started by taking a step back and strengthening our brand positioning. And how we go to market? And that's what Louis talked about this morning. That's what we introduced to you today. This platform first story. No one else can talk about this. No other competitor has this strength of the spinal AI ability, offering connected trusted experience, individualized experience from CX, customer experience to EX, employee experience. While driving superior business outcome, no one else can talk about this. This is purely anchored. This message is purely anchoring our uniqueness and differentiation, and we're going to push hard on this. We're already pushing hard on this. And that's the other part, the amplify. So how do we push this message out there, so more people hear about this? Of course, PR, analyst reports, awareness campaign, SCO, and name it, and I'll dive into some of them. On PR, we've been investing a lot in PR. And you can see it here that we've made huge improvement. So we went from 26% of articles about Coveo, so that the pie here is the number of articles about our performing competitors and Coveo. So we have 26% of the share of voice in that group a year ago, not 2 years ago, but a year ago. And now we're at 36%. So we're growing. More people are talking about us, but we want more. It's -- there's a lot of noise in this market, but we can do it, and we're working on different strategies to do that. Bold [ pop ] leadership. We need to be bold and Louis this morning used some bold statement that we're doing and that works. This has a lot of traction in the market. PR moment, let's amplify everything we're doing. Innovation. The launch of Gen AI is a good example that you'll see after how much it had an impact on how people are looking at us, and organic traffic on our website went crazy. So it does work. narrative shift. How do we talk about it differently? How do we bring new angles to e-commerce, customer service, search, search is back, search is cool, like Laurent talked about it. It's not search anymore. We talk about intent box where everything is happening there. Customer success, media want to hear about our customers, and you will hear about Deltek today. And we have other amazing great stories, and that's what they want to hear. They don't want to just promote Covelo. They want to promote Coveo with proven results, and we have them we just have to push it even more in the market. And awareness campaign. You may have seen this campaign in Wall Street Journal. We started this mid-September, and it's still on, and we actually extended it. We started in the U.S. We extended it in Canada 3 weeks ago. So we went big and bold with Gen AI. We wanted to have eyeballs on our brand and on Gen AI. So we invested big here to make sure more people would see that we exist and what we do. So we had big banners, [indiscernible] video . We had audio spot in their podcast. We're sponsoring their newsletter, 2 type of newsletter. We have integrated content in their newsletter. So far, those, as have been seen by 22 million -- or actually 20 million were up close to 22 million now time impressions. So this is big for us. More people are seeing this and then more people are clicking to discover what we do. The other part of what we do to increase our awareness is SEO. Our prospects are searching online before talking to us more and more. So we have to come up in their search results. So we invested in the last 2 years heavily in improving our SEO, and you can see the results. Coveo show up some Page 1, and of the -- our prospect and customer search much more than in the past, huge improvement, plus 600% in 2022 and plus 74% in 2023. And then showing up on the top 5 ranking on that Page 1, also growing really fast. This is really important because people are searching and they have to find us quickly, faster than our competition. The next one here is about organic growth. We've been growing organic growth on our website for the last 3 years to even more than that. But you can see here the big change, a big break in March when we launched our relevant [indiscernible] mentoring. So only organic traffic were not. We invested in SEO to win a lot of keywords around this, and it's worked. Plus 88% of traffic on our website, organic traffic, not paid traffic, organic traffic. People searching and then coming to us. And the other part is also interesting. I talked about that last year at this same meeting, [indiscernible] . What are [indiscernible] , people that want to meet with us. People are not asking for a meeting with us. So people coming to our website and clicking on talk to an expert, contact us. I want a demo, plus 72% last year, first half and this year, plus 60%. So more people are interested and actually come to our site and ask to meet with us. And our rule internally is that we have 2 hours to get back to them and book a meeting. The other part is analyst report. This is part of the prospect buying journey, they look at this. This has a lot of influence on their decision-making. So of course, as you know, we're a leader in the Magic Quadrant for Insight engine for the last 7 years, Forrester Wave, Cognitive Search for the last 5 consecutive reports. And now the new one we had in the last quarter is IDC knowledge discovery software, we're a leader. We're #1 in enterprise search with software reviews. So this is big, and we're investing into this. We're investing in those reports because they do move the needle when people are looking to buy a product like us. So last but not least, the most important thing that I'm doing is building pipeline, and it's growing. As you can see, with everything we're doing, we can see last quarter, we were plus 85% versus same time last year. So huge growth, use interest from what we do. And the pipeline driver are everything that I mentioned about awareness, of course, partnership, events, events trade shows are still big for us. They work really well for us. account-based marketing and Gen AI. When we put Gen AI on an advertising, we see plus 40% click-through rate. So what do we do? We put Gen AI everywhere because it works. So that's it. So now I'm going to pass it to Tom for the other 4 growth pillar.

Thank you, Sheila. So now I'm going to talk about how we're different. And I'm going to talk about our go-to-market strategy versus our competitors, and how that reinvigorates and reaccelerate our bookings and our revenue. And I'll get into some details on that. So when we talk about differentiation, everything revolves around our platform. You saw a little bit about it. You saw some of it demonstrated. But when we're in the field, everything revolves around that platform. And the 4 components that I talk about today have a significant competitive impact against every one of these competitors here. And I can tell you some stories about that and a little bit of what we see in the field. We see first point search and recommendation tools out there that aren't advanced and as advanced as Coveo. We run into those all the time. We see the platform vendors which are dipping their toes into the AI ocean and do a lot of things. But obviously, in this day and age, need to have a story around AI and even generative AI. And then we've got the build it yourself vendors. Every one of these, when I'm talking about the stories and the competitive win stories that we'll hit on have been touched. So every one of these is impacted or you can see within this presentation. So first and foremost, purpose-built AI models and true one-to-one personalization, we can show this in every demonstration in every POC in every one-to-one bake-off that's A/B tested and in production. And what we see is the ability for Coveo to handle millions of different customers or folks that are using an application with billions of different experiences, and provide an experience that looks and feels like it was custom built just for them. And what we're seeing from our competitors is I've been in search for a long, long time, old technology called Personas. So what they attempt to do is build 8 to 12 personas, if you will, and try to bucket those 1 million different customers in there. So they're not getting an experience it looks like it was custom made for them. They're getting an experience that looks like it was made for one of those 12 personas. And what that does is not deliver the type of ROI that we can deliver. And that's really, really important. This is not getting as much press and might not seem as exciting. You heard a little bit more about it from Louis and Laurent, but its every bit is important and its every bit is big of a competitive advantage. So our ability to unify that content with our unified index is a huge differentiator. And when you peel back the layers of the onion in the field, in a POC or in a one-to-one bake-off, what you see from our competitors is a ton of manual scripting, whether that be an SQL script, an ETL script, you see that. And that gets exposed very, very quickly. And once again, it doesn't allow them to deliver the type of return that we can. Next to that platform breadth. When you hear me talk about the different use cases today or the different competitive wins in these examples, every one of the LOBs that we have and every one of our use cases will be represented. And we can deliver that spinal AI across all of the use cases that Coveo handles. And then the architecture. I just want to reiterate how powerful something that Louis said was, 20 releases a day, no downtime in a completely secure environment that is massively scalable. What that allows us to do is to scale this globally and to do it securely, and to make sure that we can deliver this to as many customers as we possibly can. So that architecture is huge. And when you subscribe to Coveo, it really is a subscription to innovation. So constantly innovating and constantly being updated. And we'll talk about how that played true in some bake-offs when we talk about this. So I'm sure you've seen a ton of ROI slides from a lot of different companies. And I'll tell you how this is different because that's our recurring theme here. The difference between an ROI slide like this and the next one that we'll show you is that this is compared directly to our competitors and our ROIs are significantly greater. I'm going to tell you a story in a second. But at Caleres, 24% increase in conversion rate, Hearts On Fire, 49% increase in average order value, which is big and in Acuity Brands, 80% decrease in the searches that didn't bring back results. And those are all compared against competitors and were all competitive wins. But I'll tell you a story about a deal that we closed last quarter at a $10 billion retailer, I'm from Chicago. I'm in the Midwest. I know this retailer well over the last 30 years of being in sales, I've sold on 3 times. I can tell you that they're one of the toughest customers in the world to close. So we did it. And the reason we did it is because these folks did an A/B tested competitive production pilot bake-off for their customers between Coveo who got ready in a week and our competitor who had 2 years to optimize. So we increased by 24.8% in terms of conversion. Average order value increased by 9.3%, RPV or revenue per visitor, plus 128%. In the field, this is the #1 KPI. This is the #1 thing when we're selling e-commerce that they care about is revenue per visitor, went up 128% and the transaction uplift, so for each transaction increased by 109%. What this equates to in real dollars is in just 5 months, we had $268 million of incremental revenue. So when Laurent and I and the sales team are out there and we are trying to expand this, you can see that we have a tremendous amount of leverage, which allowed us to do a very sizable transaction in Q2. So we did that transaction, and this is a really conservative company who chose to stay on the sidelines a little bit for the generative AI journey. So let me just see what goes on there. There's a lot of noise in this. There's a lot going on. 24 hours after that transaction, they reached back out to us and we did a significant generative AI deal, and they're trying to go to production right now, and we're helping them with that. So that is a real life story about how all the things I just talked about on that previous slide equate to production ROIs that we can prove out. Same type of slide, different line of business. Again, super important outcomes. In AI, as we've known it, up to this point is all about outcomes, as Louis said. So in Informatica, 3.5 million an explicit case deflection, kind of similar to we were talking about a little bit later. Earlier with Xero, Formica, a 27% improvement in CSAT. And then at Xero, as you heard about from Laurent, 37% reduction in cost to serve. That's a heck of an outcome. And is, again, what we're made to do or what we've been made to do for the last 10 years, is feed a tremendous amount of data into our purpose-built AI models and deliver these types of outcomes for our customers. The difference now is that for the first time in history, AI versus Gen AI. With generative AI, we're actually creating new content that is designed to solve those problems. And when you do that, you're not just creating outcomes you're allowing people to do more by themselves, and you move from a situation of efficiency to proficiency. And even more importantly, the outcome of this was a 40% reduction in the time it took for folks to find answers to their questions and a 20% reduction in the cases they had to open. So just think about it, if you're a big enterprise and you've got 5,000 call center agents and you've got 4,000 or 1,000 or 500, and now you can flip that Coveo generative switch and reduce that head count by 20%. It's a massive ROI. And it's been proven in production. So the headline on generative is that Coveo is taking customers from the experimentation phase that I hear about constantly from CIOs into production and into a situation where they can actually deliver tangible results. So that's the headline. Availability in December 2023, early access to B2B and B2C. That's like the customer, that $10 billion customer I just told you a story about. You heard about Xero and how successful that was. And then you see, based on Xero, the tangible results that we can deliver. So when you think about key partnerships, I can tell you from a competitive perspective, there's a lot of things happening right now. We're going to see one of our competitors reduced their North American head count by 50% and completely do away with their European office in the next 2 weeks. The reason we know that is because they all want to work for Coveo. Unfortunately, there's not a lot of quality that we need from this particular vendor. My point in bringing that up is they would jump through any hoop for the partnerships and the relationships that we have like companies like SAP, who you'll hear from a little bit later, and for what we call GSIs, global system integrators like Deloitte, Accenture and Capgemini. The combination of these 2 things, what it allows us to do is create a force multiplier and surround those accounts and surround those opportunities with really trusted advisers, including Coveo. And the reason both the ISVs and the GSIs love Coveo is because they see us as an annuity, not only in the initial land, which we're really good at from a competitive perspective, do we protect their installed base. But they know that for every dollar of Coveo that's spent, there's 10 to 20x spent for those ISVs, and that's been proven by them, not by us. So we're in annuity. We're trusted with a gift that keeps on giving, and we give them a seat at the AI and generative AI table that is really sought after right now. Probably the most high-profile projects within the C2 suite, our AI and generative AI projects right now, and we get them that seat and we get them a tremendous amount of credibility and a great partner. So that's why it's no surprise when you look here that, that's grown by 80%, and we love our partnerships, and they're very, very, very important to us. We continue to expand outside of North America, we've tripled the size of our European team with top players there. And that's why we've closed deals like Roche and Nestle and Rolex and Philips. And as of this morning on my phone, Siemens. That's going to be a story that expands maybe even next quarter, but extremely impressive wins there, 30% year-over-year growth there in terms of pipeline, also thanks to my friend, Sheila here. In terms of Australia and New Zealand, this is a situation where we're getting pulled into deals without even having a team, with some of the most respected brands in that region, if you're familiar with it, Woolworths, the country, actually, Australian Taxation Office, Xero, which is in New Zealand. And if anyone is familiar with the region, Wesfarmers, which is a big conglomerate. So 85% year-over-year without even having anybody in country, so we're going to build that out, and that's going to scale like crazy. So the land-and-expand strategy at work. I'll talk about some statistics after I provide these examples. The first one, this is one of the biggest -- it depends on what you look at in some Gartner Quadrants or Forrester Waves, they're the biggest, but one of the biggest software companies in the world. And this is going to be 3 different use cases that we landed on. So this one was one of their product lines where we completely blew away the metrics for everybody else. And because of that, they deployed us system-wide across the enterprise and in product. So it went from a nice sized land to more than $3 million. This health care company does $62 billion online. In this particular case, we landed with a website, still very important, but not viewed with the same level of urgency as their B2B site. In this particular situation, Louis Tetu was at dinner with their CIO, and he said, "Listen, we've done this 3 times. We've tried it 3 times, the custom catalog complexity that we have, you guys have never seen anything like this. I can guarantee it, and we're going to break your software" and Louis said, "not going to break our software. We're going to deliver on this. You have my word" we delivered on it. So 1 million-plus deal, there's 3 million in my pipeline right now for additional deals out there, massive company. In the retail side, this is not the same one that I talked about earlier. This was an Internet land. So the Internet for their employees to get what they need and specific things and then landed with a massive B2C deal. We've got about $6 million in pipeline and expand for this one, big international brand as well. So with that, I think we're going to take a break. Thanks for your time. Much appreciated.

All right. Folks will meet you back here in 15 minutes, and we'll kick off with our panel discussion. Thank you so much. [Break]

All right. Maybe we can take our seats, please. Thank you. So we're going to start with our panel discussion. We have Suzanne Krpata from SAP and of course, Prithvi Mulchandani from Deltek. And it's gonna be moderated by our very own Louis. Louis?

Well. So good morning to both. So I'd like to make a quick introduction. I'm going to start with you, Suzanne. So thank you both for joining us today, making the trip -- special trip. So we have a special relationship with both organizations, and we're going to talk about that. So Suzanne is Chief Operating Officer for SAP CX globally, very important role at -- and so you have a very busy schedule. We appreciate again, you being here. I don't think SAP needs any introduction, 115,000 people, software organization from Germany, probably the best CFO/CIO calling card in the world. And specifically, as part of SAP CX customer experience essentially is everything related to commerce customer service and so on. So we'll talk some more about that. I'll just say that here in Canada, we have a little company named Shopify in commerce that we're very proud of. And of course, it's a very important company in commerce. But to put things in perspective, SAP in commerce from a gross merchandise value perspective through its customers probably processes 8x -- 8 to 10x more gross merchandise value, right?

Absolutely. So with SAP Commerce, we process over $1 trillion annually of gross merchandise value through our commerce engine. And we do that because we have the privilege to work with some of the largest both B2C and B2B organizations globally.

That's right. And so we're going to talk about that. And Prithvi is a Vice -- Senior Vice President of IT for an organization named Deltek. Deltek is a very large technology company. Maybe you could describe what Deltek does for those of -- those in the audience who don't -- wouldn't know maybe the company.

Yes. Sure. And thanks for having me.

Do we need -- we need an extra microphone for him.

Yes. So Deltek, we are an enterprise software and information solutions provider to over 30,000 a project-focused businesses around the world. So we have deep industry expertise in a few verticals. Government contracting is a very big one for us. Architecture, engineering, construction is another huge vertical for us. And then marketing and creative agencies as well. So yes, so...

How many customers do you have?

Over 30,000. Yes. So what we do for these customers in these verticals, these are all project-focused businesses. And we help them with their entire project life cycle. So we help them identify and win new business. We help them deliver successful projects. We help them manage and attract and retain talent, and then we help them measure their performance. So...

Very good. And Deltek has been Coveo customer really all along the journey for almost 15 years.

Yes. I know we have -- early 2010, we started -- I think, 2012 is when we started with your [indiscernible] Solutions.

That's right. So a dozen years ago and really evolved. So could you describe a little bit your Coveo journey for the audience to understand what we've been doing together so far?

Yes, sure. So it's -- for us, it's primarily around the customer service function at Deltek. So with our 30,000 customers who are using our ERP solutions, our CRM solutions, we use Coveo on our customer support portal, so they can self-serve and this is really important to us. We want them to be able to get answers to their questions about how to use our software when they need it very quickly without having to open up a case. And that's really important to us just in terms of making sure that our customers get the value out of our products and that they are highly loyal and they have high satisfaction and they renew and we retain customers. So it's critical on the self-service side. And then we also use it for our agents as well. So we do see a lot of case deflection, which is great. We love case deflection. But when we do get cases, these are typically pretty complex Coveo cases about how do I perform a certain function in the ERP solution. And it often takes our agents tens of minutes or even hours, sometimes to resolve these cases. And Coveo is absolutely essential for that use case as well and helping the agents get the right information they need from all of our different knowledge repositories so they can resolve the cases much faster.

Yes. So it's been a huge success. Suzanne, so as the Coveo and SAP relationship is very special. We're very privileged to be an endorsed partner of SAP, which only a few handful of applications across the world have the privilege to be, which essentially means that we're really joined at the hips that our sales teams work together in the field have common incentives with customers and so on. SAP is also a customer of Coveo. We -- similar application here, we help with all the personalization of service globally. So why is it important that SAP is both a customer and that we're also an endorsed partner for you?

Absolutely. So we actually became a customer first, and we used our experience as a customer really to take Coveo on a bit of a test run to understand the value of the solution that could actually bring to our end customers. So we use Coveo within our search function to support over 9,200 service agents globally that provide our enterprise support to all of our global customers. And so we found the benefits that we leverage from things like the semantic search as well as the AI answering as part of the service process. We identified that, that would be tremendously beneficial also to our customers within the commerce space. So based on the proven success we had as a customer, we then moved Coveo to an endorsed partner status. And when we elevate a company to an endorsed partner status, it's a pretty rigorous certification that we take this company through because when we promote somebody as an endorsed partner, we're putting SAP seal of approval on this organization to say this is who SAP recommends for this specific functional footprint. So for semantic search and for AI recommendations within customer experience, Coveo is the only organization that we work with and we recommend. And again, their success with us as a customer is what helps move that process to them becoming one of our handful of endorsed partners within the customer experience space.

And how do you think -- obviously, we complement each other so well, and we've already had quite a bit of success in the field with many large organizations. Tom mentioned a few already earlier. If you think about the SAP road map and the Coveo platform, how do you -- how important is AI and now generative AI in customer experiences for SAP in the area of service and commerce?

It's tremendously important. And SAP standards around AI aligned so well with Coveo standards for AI. So of our 3 pillars of AI, one of the things we really focus on is that we're delivering AI that's responsible. And a foundational part of that is how secure is the AI solutions that we're providing. And so again, as part of the certification process Coveo went through, we did an in-depth, you might argue too in-depth, but we'll agree an in-depth technical audit to make sure that we confirm that the security associated with the Coveo solution actually met the SAP standards. And what we're seeing for customers as they're on their AI journey, it's not a one-size-fits-all solution. And so it's really important that we can bring customers very specific and very tailored use cases like we do with Coveo, to help them meet real business needs in the moment with the AI capabilities.

Can you -- and I know we talked about the scale of SAP and so on, but give us a sense of -- just for the audience to understand, give us a sense of with or without naming these customers, give us a sense of the scale and the magnitude of the operations that they run. If you think about global commerce. If you think about -- you acquired a company a number of years ago named the Hybris Software that became SAP Commerce and now Commerce Cloud and et cetera. But we know, obviously, the scale of this and so on, but give us examples of the customers we deal with and the types of problems that we're working together on so that everyone can appreciate the scale.

Yes, absolutely. So some of our joint customers, if I just reference our name, for example, Dow Chemical. So we work together with Coveo for Dow Chemical because over 2/3 of the people that come to the Coveo -- excuse me, with the Dow Chemical site, do a search, which leverages Coveo capabilities to optimize the search. If you look at a company like Philips in Europe, they're choosing to work with SAP and Coveo together to help them improve their Net Promoter Score. And then if we come back to somebody's state side, we're working with one of the largest beer, wine and spirits distributors in the nation. They use SAP's Commerce platform to drive over $3 billion of ordering every year through our commerce platform and again, integrated with Coveo for both semantic search and AI recommendations. So this beer line distributor is present in 48 states and has a portfolio literally of millions of SKUs. But due to regulations, those SKUs are also specific to the individual states where they can be sold. So once again, this is a company that is relying on us together to generate billions of dollars of revenue through our joint solution offering.

And I guess since we're in Toronto, I will say that one of the key examples that we brought to this customer was actually our implementation. If you go to LCBO, you will actually find Coveo providing the intelligence behind the scenes here in Ontario. And that was a key example for that. But it's really the scale -- the size and scale of these customers and SAP in particular, right, to my right that you run commerce for more than 3,000 of the largest corporations in the world.

In the world, correct.

Yes, and commerce across the world for those organizations. So needless to say, this is a very important partnership for Coveo and we'll get back on that. Back to Coveo at Deltek, can you share some of the some of the early results that -- not early, but some of the results that we got over the years, you already talked about that, but I don't know to what extent you can quantify. And then the follow-on is as we embark on generative AI because we got into an agreement recently to deploy on top of Coveo at Deltek generative AI. And so if you could talk about, Prithvi, the -- how do you think about the metrics and why are those metrics financially important? And what do you expect more from generative AI? How important is that for Deltek?

Sure. Yes. So just a little bit about our journey. And I'll say, Louis, that we've been customers for 12 years, which is really remarkable. We've evolved and changed and completely transformed our business. And in my role as VP of Business Applications, it's incredible that Coveo has evolved with us and continue to meet our evolving needs. It's not something I see with all of our third-party vendor solutions where they tend to have a shelf life of 5 years, and then you kind of outgrow you outgrow the solution. So I'm really excited that 12 years later, Coveo is really helping fuel our growth and keeping up with our needs. We've had a lot of success over the past 12 years. I talked about case deflection briefly. And depending on how you measure it, it's -- we're deflecting about 1 out of every 2 cases. So that is huge. So we have 300 support agents. So if we didn't have a solution that was working as well as it is, think about what that would mean in terms of how big our support team would have to be-- It would be a 600 person team.

You will have to double the team.

Just not at all something we were interested in doing. So that's been a big one for us. We do customer loyalty survey, CSAT surveys. And we hear often that as we kind of -- as we've evolved Coveo and we've added some of the AI capabilities, et cetera, and new data sources, we get positive feedback from our customers on how they can find the answers to their questions without opening up a case, which is important because these are a customer trying to do month-end close and they are under a tight deadline or they're trying to process payroll. So it's really important for them to be able to quickly log in, get the answer and move on rather than opening up a case, waiting for a response. And the other success I'll point to, not necessarily a metric, but the way that we've grown, a large part of our growth strategy has been through acquisitions, and we will acquire often smaller software companies that have products that are adjacent to our ERP solutions, for example. And it's not uncommon where the companies we're acquiring have a lower renewal and customer retention rate. And we oftentimes, one of the value drivers for our acquisitions is we think we can raise those renewal rates, and that's why we acquired some of these companies. And the way we do that is we very quickly onboard them onto our support tool set, of which Coveo is a huge part. And it's been gratifying. I've seen this time and time again with multiple acquisitions over the years where we've been successful in raising those retention and renewal rates in CSAT scores. So that's another important success point for us.

So making acquisitions accretive should become part of our value proposition.

Absolutely. Yes, that's a big one.

That's a bog one. No, it's true. We have many, many customers who actually throw our indexers and move acquisitions into the mix very, very quickly. And how do you think about now generative AI? Because we we're now working together. We're in the early days of that, obviously, in testing that and so on. What are some of the challenges that you can talk about openly about deploying this, and what are some of your expectations, in fact, on how this will create even much more value as we saw earlier this morning, maybe?

Yes. Yes, sure. So yes, in terms of the value we're expecting, it's huge. I mean, we think this will be a game changer for us. So what we have today with Coveo traditional AI that we've had for a few years now, it's very important for some of the reasons I discussed. But when you think about it, an agent is still typing in a query into the traditional AI. They're seeing a list of results. They have to clicked through each of the results, process the information, synthesize it and then compose a response back to the customer. And with Gen AI, we think there's a huge opportunity to automate several of those steps I just described. So we are very excited about another huge leap in agent productivity. We're starting at -- like you said, Louis, it's just only been a couple of weeks with the early pilot. So we got to go through a few rounds of fine-tuning here. We think it will probably take us 2 or 3 runs to get to where we need to be in terms of case resolution rates. But as soon as we're there, and we're comfortable with that for an internal employee audience, we would love to put this out on our support center as well, and then we'll start seeing some case deflection benefits, too. So very excited about Gen AI. I would think it will be kind of a quantum leap from what we already have, which is pretty good.

Yes. So for sure, we'll get back on that. I want to get back to you, Suzanne, and talk about AI and generative AI. Obviously, as we made in the comment earlier opening this meeting here, it took ChatGPT to wake up the world, take the world by storm and wake up the world to AI. And obviously, people now understand AI is a different paradigm and so on, obviously, it's been around for a little longer time and so on. But you guys at SAP talk to the largest companies in the world. Arguably, you're the most important software -- enterprise software company in the world.

I would think so, but I may become bias for that perspective.

Yes. But certainly, in the ERP business, you're the GBP 800 million or less. So that's quite a privilege to talk to all these companies and so on. Could you comment on the importance of artificial intelligence and now Gen AI? But AI in general, AI and Gen AI together. And what's different this year from, let's say, last year and 3 years ago from what you're hearing from customers as you talk to the C-suite and customers across all kinds of industries, right? You guys are in manufacturing, distribution, retail, banking, you're everywhere. So what's the difference? 3 years ago, 1 year ago and now.

And now. I think there's a couple of things that are different. So first of all, I think this year, especially AI is no longer the pet project, right? It's no longer just let's do this over here and see if it works. The executives that we're talking to understand that AI is going to fundamentally change how they do business. And so it's no longer a maybe we'll do it. It's a when and how are we going to do it and make sure that it's successful. So it's really become a business imperative rather than just a business option. The second thing that the customers that we have the privilege to work with, I think understand that one of the most critical components of the AI is the fundamental data, right? AI is data-hungry. And one of the things that we have the privilege to do is work with enterprise organizations across their entire portfolio of operations. And so what that means is while I personally focus on customer experience, we all know that sometimes negative customer experiences don't happen at the point of experience. They actually happen at the point beyond the experience, right? When the order is delayed, or the shipments delayed, or the manufacturing run doesn't occur as planned or maybe that there's product damages. And so what I've been really excited about is in the conversations we're now having with executives is they understand that working together with SAP, we can bring them a unique perspective because we can absolutely address individual use cases. But what's really exciting is when we really address those enterprise use cases and we bring the foundation of all of that data that's in the ASP, ERP backbone together and really let AI go after all of the wealth of data to improve the ultimate business processes.

Yes. I think this whole idea that you mentioned, Suzanne, you said it so right, AI at the point of experience. And that's really the critical part here is there's a lot of -- you can put AI into data analytics. You can put AI into content creation and all that. AI in real time at the point of experience is going to change the world, and it sounds like your customers would agree with that. So if you think about our partnership, we've seen a major shift as a vendor in the industry dealing with many enterprise of the large enterprise software vendors. But SAP, in particular, has really taken -- has really changed a lot as an organization towards partners like us over the past 3 to 5 years, I would say. How do you think about our partnership? First of all, how do you feel the relationship is going so far? So tell it as it is. And what do you think why wouldn't SAP build something like we do or other -- some of the other partners -- endorsed partners do? And why do you go down the route of really joining at the hips with companies like us and working together? What's changed?

Absolutely. So with an engineering organization like ours, it is always a build-by-partner type of discussion. And as you can imagine, there's always interest in building. But from a speed to market and time to value, there are strategic times where partnerships make more sense. And then when we look to partner, we look to partner with best-in-class because once again, we're putting that SAP endorsement on the company that we're working with. And so with Coveo, again, based on our experiences, it was easy to see where your organization was truly best-in-class in this space. I think the other shift that I've seen within SAP in the past 3 years is our ability to continue to focus on broadening the ecosystem because not everybody owns everything from SAP. They have a robust portfolio. And sometimes there's even relationships that are already established. And so I've definitely seen a change within SAP, again, to make sure that we're coming to market with an open approach, with a combination of SAP solutions and in this case, Partner Solutions from Coveo. So now as we talk about how is the partnership going. Well, I don't really want to answer that. No, I'm kidding. I had to say that. The partnership is going extremely well. What we have found when we work with different organizations is there's multiple different ways for us to engage in what I would say our teams have done very, very well together, is engage upfront and really defining what is our go-to-market strategy, who are we going to go after together with our joint solution and our joint value proposition and then we've executed on that plan. And like any plan, the most important part of the plan is mutual accountability. And what we have found is just tremendous support from our organization about holding both of our teams mutually accountable for executing on some of those great wins that we have talked about. So the partnership is going well. Actually, within our CX space this year, it's the fastest-growing endorsed partnership that we have this year. Don't worry our budgets will reflect that next year, of course. But the partnership is growing and growing very successfully for both of our organizations.

Mutual. Same experience. And maybe in closing on that, if you're an SAP seller, if you're -- or if you're an SAP account executive, or account manager in charge of your customer and so on. In a nutshell, from their vantage point, what competitive -- what's in it for them and what's in it for the customer? What competitive advantage does that provide as opposed to ignoring a Coveo and going to the customer with already a very rich portfolio of SAP?

So one of the things that really helps our sellers is in this model, we take the sales cycle together as a partner. So the Coveo sales team is an actively engaged team member in the customer sales cycle. And so for our account executives who have a very broad portfolio, it helps to have a subject matter expert in this space, walking with them through every step in the customer journey. So for our account executive, it's access to really that specialized information, both for themselves as well as for the customer. And then from a compensation perspective, our account executives are compensated on the Coveo Solutions, just like any other solution from SAP. And in fact, with our endorsed partners, our account executives actually received accelerators on those solutions. So there's also financial benefits for our account executives to work together. But at the end of the day, the feedback that we get from the field is when we work together with these partners. It's the organizations that choose to have their sales teams walk right along our sales teams together that make the partnerships most effective.

Yes. That's right. And so we're very, very pleased with the partnership. And obviously, this is a major area of investment for us. So thank you so much. I'll turn it over back to you, Prithvi, for maybe final words of wisdom. If you think about AI and the evolution that we've been through, but that you also foresee, what advice or a key takeaway would you give to IT professionals and enterprises as they're exploring AI solutions? I mean you're obviously at the forefront because you're beyond experimentation. You're now moving forward and deploying, you're an early adopter essentially. I think a lot of companies are struggling right now with that. So what advice would you give and if you think about new offerings such as generative AI and so on, how do you cut through the clutter and move to execution?

Yes. I guess I would -- a couple of pieces of advice I would have would be, number one is get off the sidelines. Don't wait for the next GA-ready third-party offering. I think this -- with Gen AI, it's proving to be fairly easy to roll up your sleeves and experiment at low cost with OpenAI, et cetera. So spin up the POCs, experiment, fail fast, learn, keep trying until you get it right, participate in EA programs like we are doing with Coveo. So that would be a big one. And then I would also -- coupled with that, I would say, I've never met an IT organization where they didn't have too much work in too few people. So what we're doing that I think we're doing well with this as we've kind of proactively carved out resource capacity to work on these kinds of POCs and early adopter programs even before we know what they are. So that -- it's changing so fast. It's like every week, every day, there's a new opportunity or new use case coming from different parts of the business. So it's nice that we have this carved out capacity available to go after these opportunities. So that would probably be my big piece of advice for IT professionals. The other thing I'd say, again, I think Deltek is doing this well with Gen AI is we've got the very strong executive sponsorship internally. And it's really our CEO that's driving this as a key strategic initiative for the company across the board. And we're not looking at this as an IT initiative. We're looking at this as a business initiative. So every line of business leader is responsible for identifying the opportunities in their part of the business and going after them and IT is there to support. So that would be my other piece of advice is the executive sponsorship and business ownership.

So I'm going to finish by asking the same question to you both. If you fast forward 3 to 5 years from now, and I'm going to use your words of wisdom point, AI at the point of experience. Can you -- think about every sector, financial services, retailers, distribution, manufacturing, health care, government, can you think 3 to 5 years from now that organizations can be competitive without using AI at the point of experience? And I don't want this to be just a leading question. I want this to be, if you think 5 years from now in 2028, will enterprises continue to thrive and compete without AI at the point of experience?

No, I don't think they'll drive and compete. And I -- one of the trends that we have seen, especially post-COVID, is we all take our B2C experiences, and we expect those types of experiences even in B2B interactions. And so I can tell you the brands that I engage with as a consumer, I'm more likely to engage in those brands when they create a personalized experience for me that's relevant. And the same thing is going to happen in B2B organizations. And again, we're all personal consumers. And so I absolutely expect and we're already seeing B2B brands both struggling and successful when they're actually personalizing that experience. And one company that comes to mind that's a joint customer of ours that does both B2B and B2C is espresso from a coffee perspective. So they do B2B like similar to K-Cup coffee pods, but they do Nespresso coffee pods, but they also do that through B2B channels in places like special events facilities, hotels, restaurants, so on and so forth. And they have taken their learnings from what they're doing in B2C for personalization. And they're now driving that into their B2B channel. And so as we look at how AI really informs and drives personalization, I think it's going to continue -- it's already important in B2C. And I think the importance will continue to grow in B2B.

Yes. No, we agree. And Nestle Nespresso is an amazing example of what we're doing together now in more than 90 countries. You have the final word, same question, fast forward 3 to 5 years. Can you imagine that any business would succeed without AI at the point of experience?

No, I cannot. Suzanne talked about kind of the customer expectations around this and personalization. 100% agree, and then I'll just talk about kind of the internal efficiency productivity, disadvantage if a business doesn't adopt Gen AI, I feel like they'd be fighting with one arm tied behind their back compared to that competitor so...

AI or die.

AI or die.

Here we go. Final words. Thank you very much, both for making the trip. Really appreciate it.

Thank you so much. I would now welcome to the stage, Brandon Nussey, our Chief Financial Officer. Brandon?

Okay. So look, you got a view of the refreshed platform messaging from Louis to kick us off. You got to see some of the interesting things that we're working on back in the labs with Laurent. You heard from Sheila and Tom, about some of our go-to-market strategies and tactics to make sure that we win this exciting market. You got to hear from SAP and one of our customers on all the great work that we're doing. So I'll wrap up with some numbers. We'll then get you on your way. We'll take some questions if there are any, and then we'll get you on your way. I'm going to start just with an overview of the business model. I know a lot of you in this room know this cold already, but for those that may not be as familiar, we sign long-term agreements with enterprise customers for subscription-based revenue, and we have strong net expansion rates. What I want you to take away from that is it's stable, it's reliable, it's foundational revenue for us that lets us really plan and forecast the business in a reliable way. Business models, especially during these economic climates are super important, and these strong foundations set us up well. Of course, we're not immune to everything that's going on in the enterprise software world right now. But with these solid foundations underneath us, it gives us a really healthy position to keep growing our business. With those in place, we reported our numbers a couple of weeks ago for our second quarter of our current financial year. We delivered on the revenue commitment for the quarter, $31 million and change of -- in U.S. dollars of total revenue. That was up 15% year-over-year. Those that follow us closely know that we're dealing with a little pocket of Qubit related churn from an acquisition that we did of that company a couple of years ago. We'll talk a bit about that through some of these slides. But that's something we're going to deal with for a few quarters, then we'll be through it. We felt the bulk of that churn in the quarter we just reported, and I'll talk a little bit about that, but growth rates that Qubit churn for the second quarter were 19% on our subscription growth, on our subscription revenue line. Net expansion revenue. So NERs, this is how much customers pay us over and above what they paid us a year ago. So 111% when we exclude that Qubit related churn, really healthy number, something we're really proud of. Adjusted operating loss of just under $1 million for the quarter. That was well ahead of what we had planned. And so we continue to track well on our profitability metrics. It was our second straight quarter of positive cash flow from operations. We made a commitment to be cash flow positive next fiscal year. We're tracking well ahead of that. We did update our annual guidance for SaaS subscription revenue, $117 million to $118 million, down a little bit from $118 million to $120 million that we had previously said. Louis mentioned some of the dynamic in the market when he kicked off the session today. Just a lot of hype in this market, a lot of noise. You heard from Deltek, even the exercise that they go through. Just to sift through the noise, make good decisions around this stuff that did delay sales processes, installed some decisions in the market. You'll see in the coming slides, lots of reasons for us to be optimistic as we look forward. But certainly, that was a dynamic we've been dealing with for this year. And as you all know, if bookings end up being more back-end weighted, you lose months of revenue recognition, and that's what caused us to update our guidance slightly. So moving on. This is our most important line item. This is our SaaS subscription revenue. This is 93% of our revenue, of our total revenue. This carries 83% gross margins. So this is the line we pay the most attention to. As you'd expect with those foundations I talked about, strong reliable growth, 15% year-over-year, 19% when excluding the Qubit related churn. So good track record of growing that. Underneath SaaS subscription revenue is our ARR. You can see on the chart on the right -- or on the left, sorry, the Qubit related component of our ARR and how that shrunk in our second quarter. About 60% of what we expect to incur that Qubit related churn happened in the second quarter. We do expect a little bit more as we end this fiscal year, but then we'll be through it and we can keep on keeping on. ARR did grow at the rate of our SaaS subscription revenue, 19% in the core. And just a quick call out on commerce. Commerce is a relatively new area for the business from a standing start, what, 3 years ago or so? You'll see it's the fastest-growing segment, up 50% year-over-year. That's where the Qubit component is. So -- and again, for those who aren't aware, this Qubit suffers as an intentional decision by the company to not continue to invest in a small component of what they did, and that's what's driving some of that churn that we're talking about, onetime in nature. If we look at the constitution of ARR, it's a good diversified mix across the various solution areas that we serve. You can see the growth in commerce in there over the last couple of year period. All solution areas are growing greater than 10% a year, but obviously, Commerce leading the way right now. We've got strong record, a track record of delivering success for our customers. This is something we're most proud of around here. On the left is net expansion rate. So how our customers grow their spend with us over time. On the right is just the gross retention. So forget upsell, cross-sell customers paying us more. Just how much of this do we renew straight. So we see that number in the mid-90s. That's a great stat. I'm sure you're all sensitized to that. And on the left to see net expansion rates in the 110% plus is a great number for us. So really proud of this. Again, sets us up well to plan and grow as we look forward. Underneath the NER, you can see the average ACV per customer that we drive, about $175,000 per customer right now. That's up 10% year-over-year and up 25% from 3 years ago. Close to 2/3 of customers upon renewal end up spending more with us, which is a great stat. And as you'll see, we think there's a lot of opportunity to keep this going in a positive direction. This is a key growth area for us. So that's a bit about the past. Look, you've heard all the things we're working on, lots going on in our business, in our market to accelerate our growth. We think there's a lot of different areas that can come from. I'm going to focus on 3. And I'm going to -- you heard from Tom, sort of the strategies and tactics of which we're going to go about that. I'll talk a little bit about the sizing of each of these. And again, I'm picking on these 3 because they're significant undertakings for us. We've got a lot of energy around them, and they're here and now. These are the immediate drivers for us to accelerate growth from our current levels into the future here. So generative AI, obviously, our commerce business, specifically with SAP shoulder behind it, and just the ability to continue to grow the white space in our customer base. So first, generative AI. We've got close to 700 enterprise customers. Not all of them are, we think, a perfect fit for generative AI use cases. But when we go through the list, we think the majority of it is are. We think there's a good opportunity there to sell our generative offerings into the majority of our customer base. Off to a good start. This isn't even a product that's generally available yet. We're working with over 10% of our total customer base. Tend to be our largest customers as well are active with us in various stages of the pipeline right now. You've heard from us, we've signed our first 5 deals in our quarter that just ended. We signed some more into the current quarter. We expect that to continue. And when we look at the opportunity here inside the existing base, we've priced this at 40% of a customer's ACV. We've put a minimum of $150,000 on it. And again, we think the majority of our customer base, there's some application here. We're not going to obviously bat 1,000 on this, but good opportunity here for us to grow bookings, simply by bringing the excitement around GenAI to our existing customers. But generative goes beyond our existing customers. We're really proud of how fast we brought our offering to market. We think we're ahead of the market in terms of being able to deliver this in live production environments. And it's really going to help us win more customers. I mean generative is going to underpin everything we do. Nobody in this space is going to make a decision around search and recommendations around customer service, around workforce search and proficiency without understanding a company's generative capabilities. And the fact that we're there and we're there first, I think it's going to really set us up to win more. So that's a generative opportunity. If I transition now to commerce and specifically, I'll focus on SAP. We talk a lot about it. We're more privileged to have them here today. You heard about 3,000 customers in SAP's commerce ecosystem. We're just getting going. We're excited about the momentum. You heard some of the work that goes into this of enabling their sales team to do the account mapping and the planning. We've been generating this pipeline. We do expect these deals working with SAP to qualify more quickly, to close with better success rates and to be slightly larger than our average customer, given the nature of the customers that they deal with. So if you take -- if we even get to 10% of this customer base with our current ACVs, let alone whether or not we can drive higher, you can do that math quickly and calculate that's a $50 million bookings opportunity with SAP here with us as an endorsed partner. So really excited about this. This is a growth driver for fiscal '25 and beyond. These deals don't happen overnight. But you can see the momentum building, and it's an area where we're certainly very excited about. And then last but not least, is just our existing installed base. I thought this would be the right way to look at this opportunity. 70% of our customers, and we cut this off at customers that pay us $50,000 in ARR or more. So kind of eliminating the long tail of this. About 70% of our customers pay us only for one solution area right now. Those customers, on average, pay us about $160,000 a year. Once we get into a second solution area with the customer, that ACV on average grows to $350,000 a year. And if we can start to get them to adopt 3, 4 plus generative, plus other things, $650,000 to multimillions a year is what we're able to get from our customers. So when we go through our customer base, and we think conservatively about what the opportunity here is, aside from generative, which I've talked about, we see a $50 million bookings opportunity in the coming years for us. Simply by positioning, you go back to where Louis started, this is a market that we think this platform play is increasingly applicable, and we think that sets us up well to have better success in cross-selling our existing customer base. So these are the 3 areas that we're -- that I think are the reasons for optimism around how we are going to accelerate our revenue from here. The way this will play out, is again, there's no magic wand. This doesn't happen in step functions overnight. You can see the momentum building in some of these things. It's going to start with bookings, and then our revenue recognition will lag. That's the way the model works. We'll close deals, the revenue follows after you close the deal. So we expect we'll start to see that bookings momentum happened into our fiscal '25, and we expect that we'll have the revenue that follows. Of course, we are going to have to digest that Qubit churn, and that will have to work its way through the revenue line as well. And we'll have a lot more to say about this as we provide annual guidance in the coming months. But that's the dynamic at play. We think the first domino in this is closing more bookings and then you'll start to see it show up in the actual recognized revenue line. That's growth. If we transition to the other side of the ledger, the profitability, I think this company has done a wonderful job of balancing growth and profitability. I mean just -- you heard some of the initiatives from Sheila around increased publicity and some of the brand building in PR. You've heard a lot from Laurent as to some of the innovations we've been able to deliver. What I like about that top chart is the height of the bars haven't changed in 8 quarters, right? We're roughly doing all this with roughly the same amount of OpEx and you're seeing the leverage come through in the bottom chart, which is the operating loss line. We've been very disciplined about making the most important things and most important things. So when something like generative comes up, we're not a company that's needed to go hire a team of 100 engineers to pull it off. We make it the most important thing. That's where the budget goes, and you can see some of the results. So I think we've done a nice job of balancing growth and profitability. You can see we're getting darn close to adjusted operating loss neutral. I think we were actually adjusted EBITDA neutral in the last quarter, and we've been cash flow positive now for 2 quarters. So I think we're doing a good job here. There's lots of growth potential. That's obviously our priority is to grow this business. But doing so with that commitment of being cash flow positive next year and beyond. Adjusted operating loss excludes the biggest item, which is the biggest item excluded from adjusting operating loss is stock-based compensation expense. To give you a sense as to where we are here, this is something we also pay attention to. 14% of revenue in the first half of this year, down from 17% a year ago. Stock-based comp is influenced by our share price. It's influenced by historical grants. It's influenced by a lot of things. I like the middle metric myself, which is the annual dilution that goes to shareholders. That's been around 2.5%. And think that's a healthy level. And of course, on the right, we've been buying back shares. So through our SIB and our NCIB, we've retired 5 million shares or thereabouts through the first half of the year. So trying to be good stewards of capital as we go and try and win this market as well. So that will wrap me up. Just to summarize, hopefully, everything you've heard today, we think we've got an exciting platform opportunity in a market that's obviously exploding. There's a lot of excitement around this market in general, and we really like how the market is coming towards the platform that we've built over the last decade. We see a lot of growth drivers that we think we're poised to see our revenues accelerate. And the least of which is generative AI. You've got the SAP and expansion of our existing customers as well. Strong underpinnings through our reliable revenue model with strong NERs. We've got strategic partnerships that are going to help us make sure we get there. We're cash flow positive, clearly in sight, and we've got a strong balance sheet that's going to give us flexibility. So we like our position. And thank you, everyone, for spending some time with us today, and I think that will transition us into the Q&A period, and then we can get you on your way.

I Invite all the speakers for today to come to the stage. Go ahead, Thanos.

Thanos from BMO. So there was a slide showing that your pipeline is up over 80% year-over-year. Last quarter, your adjusted sales and marketing expense was up 3%. So it's great that you are cash flow positive the last couple of quarters. But just given the tremendous market opportunity, are you comfortable spending sufficiently to capitalize on it?

Brandon, do you want to take that one?

It's always a balance, right? So we're spending about 42%, I think, in the most recent quarter on sales and marketing. That feels like the right amount for us. As we see some of this pipeline turn into bookings, which will turn into revenue, the thing I feel best about is I feel really in control of the business model, like we know what we're spending, why we're spending, what our expectations are. And if we see these things work, that's a dial we can turn. We're encouraged by some of those leading indicators. I'd like to see it turn into bookings and revenue then. And as we see things that work, we can lean into them. But for now, I think we're balancing the growth and profitability well.

And maybe as a complement, Thanos, as we talked about a little earlier, we obviously want to deploy capital towards growth and want to see the growth rate of the company pick up as the market reopens. So we're in the -- we're an early player in this, and the industry also, if you look at AI and GenAI right now, the companies that monetize, we're seeing the early innings of that right now. So as soon as we see, as Brandon said, signals pick up and et cetera, obviously, we're going to capitalize on that, and we're not going to hold the expense just to reach the bottom line focus on the bottom line, it's all about growth, basically, but it's all about efficient growth. As Brandon said, it's a balance. But we certainly have capital to deploy. And if we see growth, we're going to deploy it.

And just one more for me is, can you expand on the go-to-market motion when you're trying to capitalize on cross-sell opportunity? So is it typically the case that it's a different head of the line of business that you got to work with? What challenges does that represent as you're trying to capture that? Or is every customer different in how they're set up in that regard?

You mean different, yes. You mean well, different use cases. Yes. So we're making important changes in the organization. We have, as Brandon mentioned, a big opportunity to capture winning new business. And then we have a big opportunity, obviously, growing the platform into new use cases. So as an organization, we're changing the structure of our go-to-market, global go-to-market to put much more emphasis on account management on the one hand and on growing sales and more aggressively acquiring new customers on the other hand. So from an expertise perspective, think of it as it's really at the end of the day, although the beauty here is it's one platform, multiple use cases that we can expand. At the end of the day, it boils down to 2 core sets of expertise within the company. One of them, which is what we call knowledge and customer service solutions, which is everything that deals with knowledge management, intranet self-service, customer service, it's all real search types of activities and then commerce as the other one. And our teams can overlay in doing that. So we think we're putting in place an infrastructure that can scale with that model.

Yes, all I'd add is that we've got expertise in both those lines of business and we continue to have expertise in both those lines of business. There's some that overlap more when we look at some of the logos that I talked about, like B2B manufacturing. The service component and the B2B commerce component are attached at the hip. And we're in a really good position because many of our competitors just don't respond to those bids, because we're the 800-pound drill in service, and we do a lot of great work with the custom catalogs, and we can provide that closed loop. So it doesn't provide an impediment at all. The teams communicate. And for the opportunities in the accounts that have both and many of them do, we're managing them pretty effectively.

Doug Taylor from Canaccord. A question for Laurent. I know you were discussing some of the recent developments with OpenAI over coffee, I wanted to drag that conversation out into the public here. There's been a lot of developments with OpenAI GPT plus recently, a lot of headlines. I wonder if you could just talk through, I mean, how you see the relationship with OpenAI evolving over time as the product road map -- their product road map evolves and whether you think you got to collide or intersect at certain points?

Sure. we're not going to collide with OpenAI. We are going to leverage GPT 3.5. Well, we're leveraging GPT 3.5 today, and we're going to leverage whatever large language model upgrades are going to happen in the future, just like we're going to leverage [ Cohere, Anthropic ] and all the other ones, and versions of those models that may be fine-tuned inside the enterprises for our customers. I think conversation that we were having earlier on was around security and privacy, where OpenAI is opening their system for companies, individual to push or to upload their content to have some personalized version of GPT, right? Which is in there, great. But none of our customers are interested in that. Sharing their information with OpenAI is like it's a no-go. It's not even open for discussion. Hence, why my slide about security that you saw earlier on. Very quickly, we had to build this slide because our customers are not open about this. What is going to happen, I think, is that more and more of those services will be leveraged from either Azure or AWS or GCP down the road. So this innovation that you see from OpenAI will be available on Azure, and that's totally fair. But we're going to run it on our own cloud and leverage it in our system.

I think the only thing that I'd add is, as you think about what you can do with -- or what you could do with OpenAI versus what we do, great for generative, but there are going to be -- and it's our perspective, there are going to be situations where you don't necessarily want an automatically generated answer. You want to browse. And that's where Coveo and what we've been doing for over a decade brings a lot of differentiation. And I think the word Laurent would use is nontrivial to build what we've developed over a decade plus, and that's a piece where no matter what happens with OpenAI, I mean, unless they go and build what we have, we feel like we're in a pretty good position there.

David Kwan, TD. So you guys were talking about the work that you've been doing with commerce customers in terms of helping drive more profitable revenue growth. So can you talk about where you are in terms of conversations with customers? And is that an incremental revenue opportunity for you?

Do you want to take that, Tom?

Sure. Yes, it absolutely is. So -- and it really reflects back on the specificity and the artificial intelligence that we have. If you don't have the ability to do that and identify what is more profitable and make those right decisions at scale, with AI, it's impossible. That just doesn't scale. It's impossible to do that type of thing. So when we're building our business value assessments, that's central to that. So as we go and pitch that, that becomes central when we have more and more use cases, some of which I shared with you during the presentation, that becomes even more strong of a lever to do those deals and to demand higher premiums for our software.

I know last year, when you talked about it, when there are obstacles was customers not willing to share that information with you. Have you done a change since you've kind of started talking about and rolling out the solution?

Yes, it's a journey, right? So I think it's about proving ourselves. Once we prove out with the metrics that you saw there upon, then every customer is going to be different. There are some that will never do that. And there's some that once you prove yourselves out are going to be a little bit more open to that. So software has been about really representing the value since I've been doing it for 30 years. And as you establish these relationships and as you execute, you get a lot more leeway and flexibility to recommend new things. So I think it's an ongoing journey with a lot of these commerce customers. And as we continue to execute, which we will because of the AI, then we get more opportunities open up to us.

I would just add on this one. In commerce, this is the holy grail. And the reason why customers weren't willing to share their margin data is nobody could do anything with it. When you can demonstrate tangible results on margins, we think at the end of the day, unless I'm mistaken, that's what companies care about. This audience would understand that better than most. And -- but that's extremely important. And I'll go further. I think search engines, bad search technology is actually responsible for many of the financial flows of retailers right now. They don't realize that search engines keep learning about popular products on sale and keep serving more of those, leaving the long tail of full-price products in the warehouses that then companies need to discount later on, and it drives profits down. We can actually demonstrate that at scale. And that's important. We think -- we would love to do business with Canadian Tire and help them recover profitability because we think that's a big -- that's a huge topic. And that's just an example. We don't do business today with them. I'm just picking up an example that's topical here in Toronto. But that's going to be really, really important. So the conclusion on this is it's not so much that companies aren't willing to share it. It's -- to the extent, what are you going to do with it? And can you use that data intelligently. But obviously, if you want to maximize margins, you need margin data at the end of the day. You can't do that without margin data. So they'll have to go there. Otherwise, their competitors will. And they already do, by the way.

And I think the other thing, Louis, right, it's a mindset change. And there's been such a focus in the past within the commerce space and the e-commerce space around how do I optimize the revenue. And what we're working on, and I think, obviously, being able to show customers where we've successfully done this, it's kind of an aha moment or at least we think it's going to be an aha moment where revenue per visit can't be what you're optimizing, you need to optimize margin per visit. And as we roll this out and get more customers using it, I think we'll have more data appliance, and it's going to, we think, build some momentum for what we're doing at. Go ahead, Suthan.

Suthan Sukumar from Stifel. I had a question from a go-to-market perspective. Are you guys looking at GenAI as more of an upsell add-on? Or is there an opportunity to really package up as a flagship product offering? And what are your thoughts on how cost of acquisition might change depending on that strategy?

Yes. The shift we've made from a positioning perspective is really what we call the platform first message, which is if you think about Coveo, GenAI can't work just on its own. Unless you have all the other pieces, you can't make GenAI secure, current and et cetera. But also, in the example that Laurent showed earlier about barbecues, you need GenAI to be -- to work with the other AI models that also understand relevance and understand the current products. You wouldn't want to generate an answer if you're a retailer recommending a product that you can't sell as an example. So it all needs to work together. So search, generative AI, conversations, et cetera, relevance recommendations. They're all one and the same. So Coveo is a platform, and you can activate these models essentially. So in the base platform, we offer a certain number of models. And you can think that over time, we're going to be able to sell a customer with the platform and then activate more advanced models that customers can touch and measure, and then understand the value and pay for as we activate that functionality. So it's all one and the same. Those are not separate -- there are separate SKUs from an advanced AI models perspective that can be sold separately, more as an upgrade -- not an upgrade as an add-on, basically, that keeps adding more value.

Every deal. So -- every deal is a generative deal now. There's no differentiation. There's no separate...

Okay. Okay. Great. And it's good to see that engagement on the partner and the line side that continues to scale. Can you speak a little bit about how you're investing in developing out that ecosystem? Like is there formal training that you're thinking about? Just curious what tactics you guys have at play?

Yes. So both on the sales side and on the enablement side for deployments. So a tremendous amount of investment. So we've got dedicated folks to take SAP, right? They just do SAP that are hired from SAP that understand that ecosystem front back, have relationships developed over the last 20 years. And then we've invested in folks that know how to deploy commerce and service so that those GSIs, the goal is for them to have a practice. And I can't tell you how much more credibility we had when we moved to endorse SAP. So it immediately makes the Capgemini's, the Deloitte's, the Accenture. Those practices within those organizations stand up and take notice as to what's going on and be a lot more open to investing in us. So it goes both ways. So that's both on the sales enablement side, all the way from their AEs at SAP and Salesforce, all the way to their [ SEs ], the folks that are the technical resources over there. And then on the implementation front, making sure that they know how to deploy Coveo.

Hi. I'm David Weiss with Scotiabank. So I have a couple of questions here. So you have a pre-existing solution with LLMs with Smart Snippets as I understand it. And I'm just wondering in terms of your new generative AI offering, if that would essentially supersede that, that would be potentially a sort of an upsell capability.

So for those who may not be familiar, Smart Snippet was a model that's using a large language model would extract the best paragraph or the best snippets and the best result to answer a question. Now what we're doing is we are basically taking all of those best smart snippets, so just one, all of those matching Smart Snippets that matches semantic search, and use that to ground the prompt to come up with a thorough answer in large language model. So it means that most of the time, yes, generative answering will supersede Smart Snippets. There may be specific use cases, specific customers where Smart Snippet is enough, right? Because it's a little bit faster and things like that. It's less expensive, but most of the time, yes, the answer will be better, richer, more thorough and complete.

Okay. No, that's great. And then in terms of implementation times, you had mentioned pretty rapid times for your customers on the order of weeks to months. Just was wondering if you could get a sense of how that -- if that is significantly faster than peers in your opinion?

The biggest amount of time is all of the connectivity layer, the security layer, the relevance layer, everything that's already there with Coveo, right? So -- and by the way, building that all on your own is super expensive and super long, right? So Coveo will be a lot faster. And once you have that, adding generative AI answering on top of it, is, yes, in the best case scenario, it's a few weeks, and you go into production.

That's great. And then last one for me is just in terms of analytics capabilities within your products. Obviously, you've got AB testing capabilities. Just was wondering if you could comment on the breadth and depth that's offered to customers in that. And I know I think you have provided access to Snowflake for your customers. Just was wondering if they get some.

Oh, yes, yes, yes. We have reader accounts available for our customers on our back Snowflake back end. So customers can go get the analytics data and consume it in their own dashboards and tools that they want, in addition to everything that we provide out of the box obviously.

[ Cadre ] from [ ME Capital. ] I just wanted to ask on the commerce use case in generative -- relevance generative answering versus the customer service use case. You've obviously seen that -- you were obviously seeing that you guys already have a use case live with the customer service use case. But in the -- is the heavy lifting for customer service done where you can transfer that over into the commerce use case already?

Short answer is yes. I think we figured out essentially how to, again, deliver secure, current, accurate and traceable answers cost effectively, and that's the big deal. Going back to what Nick mentioned earlier, everything is going to be generative now. That's the new paradigm, essentially is not only it's personalized, it's generative. And going back to what I said earlier in the earlier presentation, I think across all digital experiences -- people go online to do essentially 7 things. And they go online to learn something, buy something, watch something, listen to something, fix something or connect with someone. Essentially, that's what people do regardless of that use case. The big change here is that these experiences become advisory, which is how to, when should I? You can search answers, the what is or I'm looking for that or find me that content, it's retrieval, -- it's finding and retrieving and it's what it is, right, define this and find that. Generative takes you into the world, again of a think about that, an advisory experience that says, in what circumstances should I do this or describe your situation, what's your advice, what product do you recommend, how do I install a barbecue outside, what should I be thinking about if I buy this product, what about the long-term reliability of this product versus that product versus that product given the context time in. I'm going fishing up north. This is the kind of fish. This is the kind of them -- looking to pay average or no. I mean we do Bass Pro Shops. So yes, I am a candidate for $1,000 graphite fishing rod to go Salmon fishing, whatever. I'm just giving you simple examples that everybody can understand. So for us, it's really that paradigm that we enable. After that, everything is a declination of that. So Coveo becomes -- what's really interesting here, and I think the point we're trying to get across with is we've created over a dozen years, it's an extremely powerful platform with AI models that can be deployed to enable the paradigm. The fact that we've been able to take the Coveo platform into commerce already says a lot about because we're dealing with products, not personalizing information for customer service. But the reality is it's the same. If I understand your context, your intent and your behavior, and I understand whether it's products at the other end or content or both in the case of generative AI in commerce, you're blending content with products in context, which is extremely powerful. And I know this is a little conceptual, but this is really the heart of what we enable. It's hard science, but it's science that we master and it's extremely powerful, and it's extremely -- we've never seen anything that drives as much value in digital than that. And so that's a long answer. But the short answer to your question is for us to take generative in commerce leverages on what we've already done elsewhere.

And I think just to add to that, Louis, the big thing here, right? Is it's not a separate product. That's the real big point here, that it's all the same platform. It's different applications to that platform. So everything that we've done in service already in workplace as well and what we've done with generative, can now be leveraged in commerce. And that's why similar to how quickly we move from announcing generative AI to being in production, we're able to move very, very quickly from what we've already done to moving into commerce. And we mentioned on the earnings call, already a customer that signed an order form for a commerce use case. So that's very exciting.

And maybe building on that and closing on that is this is something that CIOs really care about. Because if you're a CIO of a large global company, you don't want for different indexes and 6 different search engines. And fundamentally, you can learn from every interaction to serve the next. If you can follow the user through a digital journey. Your behavior on the website tells me a lot about what you're going to buy. And then the way -- if I can get Coveo into your software as you use it, then I can gather a lot of data that I can use later on as you come for customer service. And guess what, depending on what -- ultimately, depending on what you ask in customer service, I can recommend you products to sell. And so ultimately, this is all coming together. And so it's not about a lot of companies pitch the unified customer experience. If you listen to Salesforce and they're the greatest CRM company, they pitch the Customer 360 and all that, but it's all in Salesforce. The companies we deal with have -- don't have only Salesforce. They don't have only ServiceNow. They don't have only SAP. They don't have only Adobe. They have all of the above, and they have tons of content everywhere and so on. So this is why we think there is a real category here when we talk about that spinal ability that the CIO wants that says, I want to be able to bring any piece of content to every single individual everywhere and carry that signal across the digital experience. We think it's a big category.

Richard Tse from National Bank. Great presentations today. I'm just kind of curious, like you've got great customer references. The technology is incredible. Why is this not growing at 30%? Is it the market? Is it the brand? Like help us understand that.

I think, again, if anything, my personal observation, not explanation because we're not here to give explanations on -- it is -- the results are what they are. My personal view is that Coveo was early in this market and that we're -- if anything, we're going to see the inflection point of that market very, very soon. That's certainly our view. I think we saw it, Richard, until 2018, 85% of our customers were the tech companies. The early adopters of Coveo were the big companies you know and those that I can only talk about those that we publish on our site. But I mean, it's well known that Salesforce is a client and Workday is a client, and Adobe and SAP, as you heard of the client and so on. These were the early adopters and Deltek, et cetera. These were the early adopters. They were the tech companies that understood better than most and earlier than most, more importantly, the value of AI and the imperative of using AI. I was actually surprised that in 2018, we signed a contract with Chamberlain Garage Doors, which is LiftMaster and in fact, so which sort of -- and they started talking about how they could use data from your garage door to drive a better experience and all of that. And then so that adoption curve is only nascent -- recent and nascent. And now obviously, generative AI, through another set of confusion. Now are we growing fast enough? This team that's standing in front of you is never going to tell you we're growing fast enough. We want to grow obviously faster. At the same time, if you look at the adoption of AI and name me companies right now that monetize AI. There's not a whole lot. NVIDIA does, of course. And yes, Azure Cloud and the hyperscalers that deliver OpenAI services for mostly experimentation and so on do. But very, very few companies actually do. It's starting. This market is starting. Now it's up to us now to grab that opportunity. All the signs are there that this is a market that will grow very, very fast.

Yes. Just almost a follow-up on that is crossing the chasm. And I imagine when you came up with SAP like the build versus buy decision. So GenAI is a topic discussion for all the Boards and companies out there a priority. How do you rise above the noise so they choose Coveo versus trying to build it themselves?

Well, I think Sheila talked about it at a tactical level, what we're doing in marketing. First of all, we have to invest to get our name out there. We don't want to be the world's best kept secret. So driving basic awareness. So it's 2 things, right? It's awareness and then it's differentiation. I think the reality is we're certainly betting on the fact that what we do is hard, the barriers to entry are very high. And at the end of the day, you can pitch, you can spend all the money in the world and many companies are right now, many PE firms are investing or VCs and so on. But at the end of the day, you have to check all the boxes. If you're standing in front of the CIO, you have to be able to say, like Laurent said, here's how we handle security, here's how we handle veracity, here's how we handle cost, here's how we deliver this and make it work. So that in a compliant way and all of that. And that eliminates, I would say, 99% of the company, literally, 99% of the companies that are out there. So it's getting in front. We have the advantage of being an enterprise place. So for us, there's a large number of accounts that you say, 21,000, 21,000 companies that we want to talk to. But at the same time, there are only 21,000 companies we want to talk to. And all we have to do is get in front of them, in fact, and really show them a platform that solves these problems in a way that we think nobody else solves. Your -- if you only used Salesforce, we would tell you, we'll use Einstein and GenAI within Salesforce. If you only use Microsoft -- but the customers we deal with, and these 21,000 companies, they're not like that. So they need to -- so either they build it themselves, and some of them do, by the way. We will unlikely sell to Walmart, because they have herds of data scientists and so on and they did it themselves. And frankly, if you go to Wayfair, what we do is no different than what Wayfair does. When you go on Wayfair, they built it themselves. They have 2,200 developers and data scientists working at Wayfair. We won't sell to them. But 99.9% of the companies cannot do that and can't build it themselves. And so -- and we don't see any analog for a platform like ours at this point. Are we paid to be paranoid? We are. But anybody can invest and send the rocket to Mars, but it's hard. And that's -- so that's our answer. And I think we are in a very -- we are in a position right now where -- and yours truly spends a lot of time in the field in front of customers and so on and so does this team. We talk to CIOs and et cetera, and this capability is just not something that they're seeing anywhere else right now. Can it change? Anything can change. Anybody can send a rocket to Mars, as I said. But it's hard stuff.

A quick follow-up. Just on the pipeline growth has been very strong. How do you qualify that pipeline, like the quality of that pipeline?

So we have a strong process into qualifying it. So we score it at the beginning of it, depending on ICP customer, what tech stack are they on? Are they on CP or are they on Salesforce? Is that fitting with where we actually succeed the most? So we score our pipeline with a clear score. And then we pass it on to Tom's team. It's like you really race. Marketing is bringing the pipeline, we pass it and then they analyze it as well with methodology I'll talk about, I guess, what after. So from a marketing perspective, we score it, we analyze it. We make sure that when we pass it to the sales team, it's high quality, and then the analyze there is a really a real opportunity there.

Yes, qualified pipeline. So we're seeing that grow as well that we didn't really talk about. But what we've seen compared to previous quarters is that qualified pipeline, meaning, for these quarters, what we can actually go after and close has been increasing. So that's where we apply our lead to close. And as Louis said, look at all the different types of use cases and understand that, for example, in generative, there's not a deal that somebody wasn't experimenting with something, but we qualify those up because we know that when we peel back the layers of the onion and when they actually test all the things that we already have built within our platform are really hard to build. So when that comes out and when accuracy and a lack of hallucination and security becomes to the forefront of an actual decision, that's where we think we can cross that chasm as companies become more comfortable with moving forward with this technology. The fact that we can say to our customers and to a group like you that we've been in production and we've absolutely executed on some ROIs is going to help us across that chasm.

So I think we've got some questions from the webcast audience as well. So maybe we'll take a few of those, and then it looks like lunch is out. And so we'll take a few more, and then we'll go ahead and have some lunch for folks.

Thanks, Nick. We do have a few questions from Taylor McGinnis at UBS and a question from a shareholder. So the first question from Taylor was around the risk of in-house builds and was already addressed by Louis and Laurent, and response to separate questions. The second is as follows: You talked about a lot of growth levers as we get into 2025, mentioned growing 19% excluding Qubit today. Any color on when we could see that inflection start to materialize and level of upside, we could see the growth? How do you think about those growth levers offsetting macro pressures?

Brandon, do you want to start with that?

I mean there's a lot in there, I guess, at its core, we tried to give you a feel as to the various reasons we stand here with confidence and optimism that we're going to see that reacceleration happen as we stare at pipeline in the back half of this current fiscal year, which will be the leading thing to recognizing revenue in our next fiscal year. Our pipeline is a lot stronger in the back half of this year than when -- if we were to go back and look at what it was, at the start for the first half of this year. So all of these things tell us that it's starting now. We feel like the macro pressure remains, though this is a space that companies are going out of their way to prioritize right now. We've seen companies sift through that noise and are starting to come out the other side. We're seeing it in order volume, we're seeing them in pipeline generation. So our expectations here is that we're going to start seeing bookings in the back half of this year, that carry in through next year. And as I mentioned a few times, the revenue will follow. So there's a way of answering that with no specifics. So there you go.

All right. Still from Taylor. With service still the single largest component, any color you can give on the growth of that business and how that could evolve based on the trends you're seeing in the business?

I think I'll start by qualifying it, and thank you, Taylor, for the question. I think actually, there is a major, major revival of service and more broadly knowledge management, because the 2 are linked, essentially. This is the -- what companies call this practice within their own organizations as KM or knowledge management. And essentially, what we're seeing in service is a major shift into -- there's kind of a combination of 2 things. Number one is obviously generative AI everywhere. So Laurent talked about in product, talked about issue anticipation, talked about in product, self-service and then within the agent. So that's a given, and then making sure that you carry that signal across the service organization. And so that's key. But also what you're seeing is a major shift of value towards the self-service component, as we've described. If we can show through generative AI, in particular, that self-service drives the kind of numbers that we talked about, which is 20%, it basically means in plain English is that if you sell seats of contact centers, you got to be seriously concerned about, because those seats are going to probably reduce. Probably not by the same percentage because what happens is the issues that get solved -- more issues get solved in self-service, then more complex issues on average, get into the contact center. And so it's going to require a bit of reskilling and GenAI will help their in-service. So bottom line is the net answer to the question is there's a major, major transformation in service happening right now and generative AI is at the heart of this, for sure. So it's a big -- it's -- we certainly anticipate a lot of growth there. And of course, we've already talked about commerce as well, and you're seeing the numbers and the growth we're seeing in commerce. So in commerce, it's a bit of a different -- I would put a different tact in terms of the growth of commerce. I think I foresee personally, and I think we do that retailers in particular, and then you heard Susan about B2B commerce wanting to apply the same things as B2C commerce. I think retailers, in particular, are going to realize that it's truly an area where it's AI or die. If you don't have AI, that optimizes both experiences and ultimately, revenue and margins simultaneously, which is not humanly possible without AI. You're going to compete against a retailer that does. And you're not going to survive that. So it's going to be pretty brutal and that creates a big opportunity for us because that imperative with retailers, retailers tend to be, no offense, a little more laggers in technology. But now I think they're going to realize very, very quickly the imperative here.

And I think just adding to the service question. we set all 4 lines of business our solution areas are growing double digits year-over-year on an ARR basis. Commerce is growing the fastest. I would say service is the one where we start looking forward at some of the generative opportunities that are out there for the next couple of quarters here. I think there's the potential to see some acceleration of that ARR growth by the end of fiscal '24. And then obviously, there's the [indiscernible] component.

And without prolonging too much, I would say, let's not forget workplace. I think we're seeing a major, major revival in workplace. Workplace was kind of a boring area, where it's productivity, it's soft costs, et cetera, who cares about the Intranet, right? It's the Intranet, pardon my French or doesn't work well, I'll use that term, in every company, but it's an ambient problem. It's nobody's responsibility and nobody cares. Otherwise, SharePoint wouldn't have sold that much. But the reality is that now, with the new capabilities, obviously, and the ability to augment employees, CIOs and CEOs and the C-suite is really going to be staring at a workforce and saying, well, at least a white collar workforce and saying, I could run my company with 10% or 20% less people. This is really what they're going to be staring at. And so it's creating a major -- and we're seeing it in the pipeline as well. It's creating an interest -- revived interest for workplace type of applications and knowledge management within the workplace.

Thanks, Louis. Two-pronged question from George McGreehan of BofA Securities, about GenAI. So first, can you talk about contribution margin for the AI add-on compared to the overall business? And second, among your customer base, what is the attach rate opportunity for the AI add-on?

So on the first, so far, so good. We've said that we're optimistic that we know enough about these models and our ability to control box in our pricing around these that we can keep margins in and around where we are and so far, so good to that. It's obviously very early innings, but so far, so good there. And in terms of attach rate, we did the one slide with the circles where we think as we go through and look at our customer constitution and what our customers are using us for today, the majority have the potential to adopt our generative solution. How fast that happens, how successful we are, of course, comes down to execution, but that's how we're viewing the opportunity.

All right. Thanks, Brandon. And then last question from a shareholder. Can you comment a little bit further on what you're seeing from a competitive landscape standpoint? Any changes over the last 12 months?

Go ahead, Laurent or Louis.

Well, I'm going to start. I think what we're seeing is that the fact that we've been building AI models for a dozen years. And the way we've been able to combine them and so on now serves us very, very well. In fact, there's a lot of maturity in what we do. Of course, everybody talks about AI right now. I think ice cream powders now do AI for some reason. And -- but the reality is, again, we're dealing with large enterprises, we do it at scale. There's a lot of intricacies here. When you get really surgical about this, that things we take for granted that are mature in our platform that we do that we don't see others do. If you combine the combined capabilities of behavioral, machine learning, deep learning, semantic search, the ability to handle large language models and generative AI together, we don't see anybody yet that has been able to master that combination the way we do, for the simple reason that we've been at it 12 years and our closest competitors have been at it 1.5 years. And so that speaks volume about these capabilities. In the world, our competitors come from 2 angles. The world of search, which in turn is kind of 2. So companies who want to build it themselves, and there are some of those out there, I commented earlier, who were going to build it themselves, big companies, et cetera. They'll take elastic, and then they need to build, all the connectivity layer, they need to build, all the relevance layer and all of that. We don't think that unless you hire herds, and what I mean herds of data scientists, we mean hundreds of people. Over a period of years, you can achieve that. But still, some companies do that and some companies like the Wayfair's and the Walmart's and some of the larger banks, et cetera, have started that journey a long time ago. So obviously, they have some degree of expertise. So that's one category, and it's really the elastic world. And so it tends to be the bill as Tom mentioned. The other competitors tend to be in the search area, more classic search vendors. And we have become experts and can very quickly point to their deficiencies because of their weaknesses in AI and their weaknesses in relevance and their weaknesses in semantic and so on, and point to the fact that those things really matter. So what we're seeing out there, we don't believe, objectively that we're seeing the kind of maturity of Coveo and for now. And then obviously, as you get into application specific, one could perceive us Salesforce Einstein, for instance, to be competitive to Coveo. Again, if all your content is in Salesforce and all you're using is Salesforce, yes, I mean we'll be the first one to recommend you to just use that. But that's not the use case here. That's -- Coveo is about creating that layer of software, that spinal ability, as Tom mentioned, to cut across multiple content sources and multiple apps. And so that's just not -- and I would argue that our degree of maturity and relevance has -- is really, really superior and you can demonstrate that through AB tests. Look, there are always buyers who will fly into an aircraft that's assembled by the lowest bidder. But generally speaking here, we're talking about high ROI areas where technology really matters, and companies don't want to provide smart buyers and large enterprises don't want to compromise on the value they can create. And so we feel pretty confident right now about our competitive position.

With that, thanks, everybody, for joining us for the morning. Hopefully, you found it informative. We've got some food over here. Please help yourself. We'll be -- most of us will be popping around if you'd like to catch up. But thanks so much for the time and for the support, and we'll look forward to staying in touch with you all. Thank you.